ReportProfilesJobs

77% of people who name Meta's superintelligence lab were already at Meta when it formed

Who names Meta Superintelligence Labs on their profile and when they joined Meta, then a map of the lab among Meta's 8,257 visible AI staff by unit, role, source, moves, and posted pay, on the Metix AI Platform on September 22, 2026.

Run itStage 2: the map353 API Credits to reproduceReproducing reads no profiles: every number is a count
722
of the 936 were at Meta when the lab formed in June 2025 (77.1%)
31
of the 214 who joined from outside had just left one of five AI labs (14.5%)

Meta put its existing AI teams, FAIR among them, into the lab when it announced it on June 30, 2025, so many long-time staff were always going to be in it. The profiles show how many, and who came in.

The 936 who name the lab, by where they stood when it formed

722 were at Meta when the lab formed, 44 had worked at Meta before but not then, and 170 were new to Meta.

Who is counted

Named the lab
The headline or current title says superintelligence or MSL: 936 profiles, 790 of which write superintelligence.
Employer
A current job at Meta. The Metix AI Platform treats Facebook and Meta as one employer; an earlier Meta job also includes Instagram, WhatsApp, and Oculus.
At Meta when it formed
A Meta job, not an internship, that began before June 2025 and was still running in May 2025 or later.
Outside hires
Everyone else: people new to Meta, people who worked at Meta and came back, and people who had only interned there.
Dates
Job start dates run to April 2026. Profiles lag, so the last months are undercounted. A date given only as a year reads as January.
What a count is
These are visible profiles of people who chose to name the lab: a count of profiles, not the lab's headcount. No one is named.

In brief

  1. 722 of the 936 who name the lab (77.1%) were at Meta when it formed. Among the 790 who write superintelligence, the share is 78.1%.
  2. The 170 who were new to Meta arrived in a burst from June 2025, the month the lab formed: 107 of them from June to September.
  3. 31 of the 214 outside hires (14.5%) had just left one of five AI labs, 18 of them Google DeepMind; 22 came from a university and 16 from Scale AI.
  4. Outside hires lean toward research: 50.5% of them work in research, against 30.6% of those already at Meta. 67.5% of the lab's research staff list a doctorate, against 41.0% of Meta's other research staff.

Already at Meta

Part 1, Figures 01 to 03

When the people who name the lab first joined Meta, whether the share holds under two definitions and at Microsoft, and who recorded a new role from June 2025, the month it formed.

01

The typical member first joined Meta in 2022; newcomers arrived in a burst in the summer of 2025

When the people who name the lab first joined Meta. Height is first joins per year, so periods of different lengths compare; the number on each bar is people

Before 2014: 14; Jan 2014 to Dec 2015: 26, 13 a year; Jan 2016 to Dec 2017: 75, 37 a year; Jan 2018 to Dec 2019: 128, 64 a year; Jan 2020 to Dec 2021: 190, 95 a year; Jan to Dec 2022: 114, 114 a year; Jan to Dec 2023: 24, 24 a year; Jan to Dec 2024: 118, 118 a year; Jan to May 2025: 77, 186 a year; Jun 2025: 20, 244 a year; Jul 2025: 36, 424 a year; Aug 2025: 28, 330 a year; Sep 2025: 23, 280 a year; Oct to Dec 2025: 29, 115 a year; Since Jan 2026, or no start date: 34, 103 a year

What it shows

766 people first joined Meta before June 2025, 81.8% of those who name the lab; half had first joined by the end of 2022. From June to September 2025, first joins averaged 27 a month, against 10 a month in 2024.

Method and limits

A first join is the start of a person's earliest Meta entry, internships included, so the part before the lab also holds those who left and came back or only interned (44 together). The last period runs to April 2026 and holds fewer than 10 with no start date; profiles lag, so it is probably undercounted. Earlier years count only people who are still at Meta and name the lab now, so they run low next to the recent months. A start date given only as a year reads as January, so a newcomer who wrote only 2025 counts before June 2025.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

02

77.1% were at Meta when the lab formed, and 78.1% of those who write superintelligence

Each row is 100%, split by where people stood when the lab formed

Superintelligence or MSL: 722 at Meta when it formed, 44 had worked there before, 170 new; superintelligence only: 617 of 790; worked at the company before its team formed: Meta 81.8%, Microsoft 76.3%

What it shows

Under the two definitions, the share at Meta when it formed differs by 1.0 points. Microsoft announced its own superintelligence team in November 2025; on the same measure, 76.3% of its 80 had worked at Microsoft before, against 81.8% at Meta.

Method and limits

Of the 44 who had worked at Meta but not then, 19 left and came back and 25 had only interned. Microsoft's 80 are profiles with a current Microsoft job whose headline or title says superintelligence; reads found recruiters among them, and the company names leave out subsidiaries such as LinkedIn and GitHub. A date given only as a year reads as January, so an entry that ended in 2025 without a month counts as ending before May 2025, which can move someone into the 19 who returned. Profiles stop in April 2026, which hides more of Microsoft's months since its team formed than of Meta's, so its share is likely biased upward.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

03

14.8% of those at Meta when the lab formed recorded a new Meta role from June 2025 on

The 722 at Meta when it formed, by when their newest current Meta role began. Bars are shares of the 722 on a 0 to 70% scale

Before 2025, or no date 476, Jan to Mar 2025 86, Apr to May 2025 53, Jun to Jul 2025 28, Aug 2025 27, Sep to Dec 2025 21, Since Jan 2026 31

Before June 2025

  1. Before 2025, or no date476 · 65.9%
  2. Jan to Mar 202586 · 11.9%
  3. Apr to May 202553 · 7.3%

From June 2025: 107, 14.8%

  1. Jun to Jul 202528 · 3.9%
  2. Aug 202527 · 3.7%
  3. Sep to Dec 202521 · 2.9%
  4. Since Jan 202631 · 4.3%

What it shows

Of the 107 already at Meta who recorded a new role from June 2025, 55 did so by the end of August 2025. Most of those already at Meta (65.9%) hold a newest role that began before 2025, or give no date.

Method and limits

The date is when a person's newest current Meta entry began. Many people change teams inside Meta without adding an entry, so this is a lower bound on who moved. The first row also holds entries with no start date.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

From outside

Part 2, Figures 04 and 05

Where the 214 outside hires had just worked, and how their roles differ from those already at Meta.

04

31 of the 214 outside hires had just left one of five AI labs, 18 of them Google DeepMind

The job each outside hire had just left, one that ended in 2025 or later. Each person counts once. Bars are shares of the 214 on a 0 to 50% scale

OpenAI <10, Google DeepMind 18, Anthropic 0, xAI <10, Thinking Machines Lab 0, Apple <10, Scale AI 16, Google, other than Google DeepMind 16, Microsoft <10, Amazon 10, NVIDIA <10, A university or other educational institution 22, Another employer, or none that ended in 2025 or later 101

Five AI labs: 31, 14.5%

  1. Google DeepMind18 · 8.4%
  2. OpenAI<10
  3. xAI<10
  4. Anthropic0
  5. Thinking Machines Lab0

Elsewhere

  1. A university or other educational institution22 · 10.3%
  2. Scale AI16 · 7.5%
  3. Google, other than Google DeepMind16 · 7.5%
  4. Amazon10 · 4.7%
  5. Apple<10
  6. Microsoft<10
  7. NVIDIA<10
  8. Another employer, or none that ended in 2025 or later101 · 47.2%

What it shows

Across the whole group, earlier jobs at the large technology companies are common: 133 (14.2%) once worked at Google, Google DeepMind included, 110 at Amazon, and 92 at Microsoft. Earlier jobs at OpenAI (12) and Anthropic (none) are rare. Axios reported that Meta took a large stake in Scale AI in June 2025.

Method and limits

A job counts when it was not an internship and ended in 2025 or later. Each person counts once, at the first match in the order OpenAI, Google DeepMind, Anthropic, xAI, Thinking Machines Lab, Apple, Scale AI, Google, Microsoft, Amazon, NVIDIA, then universities, where internships count because doctoral students' entries often carry the Intern label. Counts from 1 to 9 show as <10, drawn as an outline at the largest value they could hide.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

05

50.5% of outside hires work in research, against 30.6% of those already at Meta

Current function, and the share that lists a doctorate, for the two groups. Axis 0 to 60%

Research: already at Meta 30.6%, outside 50.5%; Engineering and technical: already at Meta 42.2%, outside 28.5%; Product: already at Meta 6.8%, outside 5.1%; Another function: already at Meta 9.6%, outside 8.4%; None stated: already at Meta 10.8%, outside 7.5%; Lists a doctorate: already at Meta 29.6%, outside 37.9%

What it shows

Engineering is the largest function among those already at Meta (42.2%); research is the largest among outside hires. 37.9% of outside hires list a doctorate, against 29.6% of those already at Meta.

Method and limits

Function is the Metix AI Platform's fixed value for the current job. Another function is any other stated function; 94 state none.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

The lab now

Part 3, Figures 06 to 09

Doctorates among its researchers, the groups people name, where they are, and the postings open on the day of the snapshot.

06

67.5% of the lab's research staff list a doctorate, against 41.0% of Meta's other research staff

People whose current function is Research. Scale 0 to 70%

The lab's research staff 67.5% of 329; Meta's other research staff 41.0% of 8,955

  1. The lab's research staff329 people222 · 67.5%
  2. Meta's other research staff8,955 people3,670 · 41.0%

What it shows

The lab's research staff are also more often labelled Senior (15.8% against 10.8%); at manager level or above the gap is within chance (6.7% against 9.0%).

Method and limits

The baseline is current Meta staff whose function is Research and whose headline and title name neither lab term: 8,955 profiles, including people who name only FAIR, which Meta put inside the lab. Senior covers senior, staff, and principal titles.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

07

Fewer than one in six name any of four groups inside the lab; FAIR is named most, by 6.7%

Groups named in the headline or current title; a person can name more than one. Bars are shares of the 936 on a 0 to 10% scale

FAIR 63, MSL Infra and other infrastructure 49, TBD Lab 13, Products and Applied Research <10

  1. FAIR63 · 6.7%
  2. MSL Infra and other infrastructure49 · 5.2%
  3. TBD Lab13 · 1.4%
  4. Products and Applied Research<10

What it shows

At most 134 name any of these, so at least 85% name none. Another 238 current Meta profiles name FAIR without naming the lab; people in its groups who do not write its name fall outside this count.

Method and limits

Four groups were searched: FAIR, TBD Lab, infrastructure, and Products and Applied Research. FAIR and TBD are ordinary words, so the matches were read: 10 of the 13 TBD matches are in TBD Lab. Applied research mostly matches the job title Applied Research Scientist, so Products and Applied Research counts only PAR or all three words.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

08

48.4% list the Bay Area, and 94.1% the United States

Where people list themselves. Bars are shares of the 936 on a 0 to 50% scale

San Francisco Bay Area 453, New York State 118, Washington State 84, Elsewhere in California 27, Elsewhere in the US, or no state 199, United Kingdom 19, Elsewhere, or no country 36

  1. San Francisco Bay Area453 · 48.4%
  2. New York State118 · 12.6%
  3. Washington State84 · 9.0%
  4. Elsewhere in California27 · 2.9%
  5. Elsewhere in the US, or no state199 · 21.3%
  6. United Kingdom19 · 2.0%
  7. Elsewhere, or no country36 · 3.8%

What it shows

After the Bay Area come New York State (12.6%) and Washington State (9.0%); 19 list the United Kingdom.

Method and limits

Location is where the person lists themselves. Elsewhere in the US includes profiles that give no state.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

09

30 open Meta postings name the lab, 23 of them in Menlo Park

Postings open on the day of the snapshot, by location. Bars are shares of the 30

Menlo Park 23, San Francisco 3, New York 2, Elsewhere 2

  1. Menlo Park23 · 76.7%
  2. San Francisco3 · 10.0%
  3. New York2 · 6.7%
  4. Elsewhere2 · 6.7%

What it shows

They are 1.4% of Meta's 2,104 open postings; 14 have research in the title.

Method and limits

One day of postings. All 30 were posted in September 2026 and all carry the seniority label Not Applicable, so neither says anything here. 7 of the 26 that state an experience requirement ask for 36 months or less.

Source: Metix AI Platform, jobs, 2026-09-22.

Queriespopulation.json · measures.json
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}
POST /v1/people/query · POST /v1/jobs/query · queries/measures.json
{  "note": "What is counted, all within the population in queries/population.json unless a block says otherwise. Every count is sent with size 1, so it costs 1 API Credit, or nothing when it finds no one, and reads no record.",  "first_joined": {    "edges": [      "2012-01-01",      "2014-01-01",      "2016-01-01",      "2018-01-01",      "2020-01-01",      "2022-01-01",      "2023-01-01",      "2024-01-01",      "2025-01-01"    ],    "after_cut_edges": [      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "A person's first Meta job started before a date exactly when they have an entry at a history name that started before it, so a count at each edge gives how many first joined before it, and the difference between two edges is a band of first-join dates. Bands run from the start of the record through the cut (the dates in edges) and on month by month and quarter by quarter to April 2026 (after_cut_edges). The last band holds April 2026 and anything later, and the fewer than 10 whose Meta entry has no start date. Adjacent bands merge wherever one would hold 1 to 9 people; the cut is never merged away. The second series counts only non-intern entries, so it gives each person's first Meta job that was not an internship; it runs to the cut, and everyone without such a job before the cut is one band."  },  "moves": {    "edges": [      "2025-01-01",      "2025-04-01",      "2025-06-01",      "2025-07-01",      "2025-08-01",      "2025-09-01",      "2025-10-01",      "2026-01-01",      "2026-04-01"    ],    "note": "For the people who stayed: when their latest current Meta entry started. A count of people with a current Meta entry that started on or after an edge gives how many started their newest current role since then, so a person with two current entries is placed by the newer one. Many people change teams inside Meta without adding an entry, so this shows who recorded a new role, a lower bound on who moved."  },  "sources": {    "ended_from": "2025-01-01",    "note": "Where the outside hires (everyone not at Meta when the lab formed) had just been: a job entry that ended in 2025 or later (experience.end_date gte ended_from), which also means it is no longer current. Internships do not count for companies (the intern rule in population.json), so a PhD student's summer at a lab is not where they came from; for academia they do count, because doctoral students' university entries often carry the Intern label. Rows do not overlap: each person counts at the first row in this order whose condition they meet, and the rest are other employers or none. ai_lab marks the five AI labs whose subtotal is published.",    "order": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ],        "ai_lab": true      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ],        "ai_lab": true      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ],        "ai_lab": true      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ],        "ai_lab": true      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ],        "ai_lab": true      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "scale-ai",        "label": "Scale AI",        "names": [          "Scale AI"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "academia",        "label": "A university or other educational institution",        "company_type": "Educational"      }    ],    "context": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      }    ],    "context_note": "For the whole group: anyone with a past (not current) non-intern job at the employer, at any time. These rows overlap and have no total. Names match word by word, so Google also matches Google DeepMind and Amazon matches Amazon Web Services (AWS). company.type Educational is the employer type the Platform records for universities and schools."  },  "roles": {    "functions": [      "Research",      "Engineering and Technical",      "Product"    ],    "senior": [      "Senior"    ],    "manager_up": [      "Manager",      "Head",      "Director",      "Vice President",      "President/Vice President",      "C-Level",      "Partner",      "Founder",      "Owner"    ],    "note": "current_function and current_seniority describe the current job and come from the Platform's fixed sets; either can be missing. Senior covers senior, staff, and principal individual contributors (the audit saw Staff, Senior Staff, and Principal titles labelled Senior); manager_up is Manager and every level above it; the rest of those with a stated level are Specialist or Intern. A doctorate is an education entry with education.degree eq Doctorate, finished or in progress.",    "baseline_note": "The baseline is Meta's other research staff: a current Meta entry, current_function eq Research, and no lab term in the headline or current title. It includes people who name FAIR only."  },  "teams": [    {      "id": "fair",      "label": "FAIR",      "terms": [        "FAIR"      ]    },    {      "id": "tbd-lab",      "label": "TBD Lab",      "terms": [        "TBD"      ]    },    {      "id": "infra",      "label": "MSL Infra and other infrastructure",      "terms": [        "infra",        "infrastructure"      ]    },    {      "id": "par",      "label": "Products and Applied Research",      "terms": [        "PAR",        "product applied research"      ]    }  ],  "teams_note": "Sub-team names in the headline or current title of people in the population, matched word by word; a person can name more than one. Audited: FAIR, 10 profiles naming it read in whole slices, nearly all naming it as the current team. TBD, all 13 matches read: 10 are in TBD Lab or report to it, and the rest name TBD while recruiting for it, while saying they are not in it, or while saying they have left the lab. Infra, the first 15 of 49 matches for infra or infrastructure (not a whole slice): all 15 work on infrastructure inside the lab, 11 of them naming MSL Infra or Infra at MSL. Products and Applied Research: the words applied research mostly match the job title Applied Research Scientist (most of those read) or an old team name, so the row needs PAR or all of product, applied, and research. fair_only counts current Meta profiles that name FAIR and no lab term; it was not audited outside the population.",  "places": {    "states": [      "California",      "New York",      "Washington"    ],    "bay_area": [      "Menlo Park",      "San Francisco",      "Sunnyvale",      "Palo Alto",      "Mountain View",      "Redwood City",      "Burlingame",      "Fremont",      "San Jose",      "Santa Clara",      "San Mateo",      "Foster City",      "Oakland",      "Berkeley",      "Cupertino",      "Los Altos",      "Belmont",      "San Carlos",      "Millbrae",      "San Bruno",      "Newark",      "Milpitas",      "Los Gatos",      "Campbell",      "Saratoga",      "Daly City",      "South San Francisco",      "Emeryville",      "Alameda",      "Hayward",      "Dublin",      "Pleasanton",      "Walnut Creek",      "Atherton",      "Portola Valley",      "Woodside",      "Half Moon Bay"    ],    "countries": [      "United Kingdom"    ],    "note": "The profile's own location, where people list themselves: location.country and location.state, and location.city for the Bay Area, whose cities are listed one by one and compared exactly."  },  "postings": {    "company": "Meta",    "title_terms": [      "superintelligence",      "MSL"    ],    "description_terms": [      "superintelligence labs"    ],    "research_title": "research",    "cities": [      "Menlo Park",      "San Francisco"    ],    "states": [      "New York"    ],    "experience_max_months": 36,    "posted_from": "2026-09-01",    "note": "Open Meta postings (is_open eq true) whose title names superintelligence or MSL, or whose description contains the words superintelligence and labs. All 30 read on 2026-09-22 are Meta postings whose description names Meta Superintelligence Labs. Every one carries the seniority label Not Applicable, so seniority says nothing here, and all were posted in September 2026, so the index shows the openings of the moment rather than a history. Counts of postings carry no minimum."  }}

Stage 2 · The map

A ninth of Meta's visible AI staff name the lab; more left Meta for AI labs than joined

Stage 1 asked who is new in the lab. Stage 2 places the lab among all of Meta's visible AI staff: the units people name, what they do and at what level, where they came from and where they went, and the base pay Meta's postings state.

Meta formed the lab on June 30, 2025, split it into four teams in August 2025 (TBD Lab, FAIR, Products and Applied Research, and MSL Infra), and cut more than 600 roles in three of them in October 2025, leaving TBD Lab untouched. It reported a headcount of 75,472 as of June 30, 2026.

In brief

  1. 8,257 current Meta profiles name an AI role. 62.8% name none of the nine units searched; the lab is the largest unit named, with 936, then GenAI, the former org, with 506.
  2. FAIR outside the lab is the most research-heavy unit: 69.8% research and 57.2% doctorates. AI infrastructure is 74.4% engineering, and ads has 34.4% at manager or above, against 14.2% overall.
  3. The lab is 35.1% research, against 17.8% for the rest of Meta's visible AI staff; 24.2% of the rest are titled machine learning engineer.
  4. Since June 2025, 379 people in any role left Meta for five AI labs and at most 107 joined from them, while 691 joined from Amazon and 343 from Microsoft.

The shape

Part 4, Figures 10 and 11

Which units Meta's AI staff name, where the lab sits among them, and how the visible count compares with Meta's reported headcount.

10

63% of Meta's visible AI staff name none of nine units; the lab is the largest, at 11.3%

Each person counts once, at the first unit named in the headline or current title, in the fixed order the notes give; the named units are sorted by size, the residual last. Each ribbon carries a unit's people out of the total, as tall at both ends as its count; the lab's ribbon runs on and splits into its four groups, drawn 4 times larger. Choose a unit to see its profile against Meta overall

A flow map: Meta's 8,257 visible AI staff on the left split into ten ribbons drawn to scale, the lab 11.3% and the majority that names no unit 62.8%; the lab's ribbon splits again into its four groups, drawn 4 times larger, and at least 802 of its 936 name none of them. A panel gives the chosen unit's six shares against Meta overall.

The lab's 936, drawn 4 times larger

FAIR63

MSL Infra and other infrastructure49

TBD Lab13

Products and Applied Research<10

Names none of the four≥ 802

FAIR63

MSL Infra and other infrastructure49

TBD Lab13

Products and Applied Research<10

Names none of the four≥ 802

Unit profile

Meta Superintelligence Labs

936 profiles · 11.3% of Meta's visible AI staff

  • Research2nd of 835%
  • Engineering6th of 939%
  • Manager+7th of 918%
  • Doctorate3rd of 832%
  • China-educated3rd of 820%
  • US2nd of 894%

Tick: Meta overallIts four groups can overlap; figure 07 lists them.

Meta's visible AI staff, by unit
UnitPeopleShare of allResearchEngineeringManager+DoctorateChina-educatedUS
Meta · visible AI staff8,257100.0%20%58%14%29%20%83%
Meta Superintelligence Labs93611.3%35%39%18%32%20%94%
Of the lab: FAIR63not applicablenot measurednot measurednot measurednot measurednot measurednot measured
Of the lab: MSL Infra and other infrastructure49not applicablenot measurednot measurednot measurednot measurednot measurednot measured
Of the lab: TBD Lab13not applicablenot measurednot measurednot measurednot measurednot measurednot measured
Of the lab: Products and Applied Research<10not applicablenot measurednot measurednot measurednot measurednot measurednot measured
Of the lab: Names none of the four≥ 802not applicablenot measurednot measurednot measurednot measurednot measurednot measured
GenAI, the former org5066.1%20%34%25%20%12%90%
Reality Labs, AR, VR, and wearables4004.8%17%40%20%24%4%82%
AI infrastructure3904.7%4%74%22%10%15%94%
Ads and monetization3143.8%11%52%34%25%22%91%
FAIR, outside the lab2152.6%70%16%8%57%16%66%
Ranking and recommendations1521.8%23%59%16%36%28%91%
Instagram, WhatsApp, Messenger, Threads1041.3%12%53%23%18%13%withheld
Integrity520.6%withheld31%27%withheldwithheld71%
Names none of these5,18862.8%withheld66%10%withheldwithheldwithheld

What it shows

The lab is 11.3% of Meta's visible AI staff and 35.1% research, against 19.8% overall. FAIR outside the lab is the most research-heavy unit (69.8% research, 57.2% doctorates); AI infrastructure is 74.4% engineering; ads has 34.4% at manager or above. Among the units shown, the China-educated share runs from 4.4% in Reality Labs to 27.9% in ranking.

Method and limits

Units are what people write, not Meta's org chart. The order is the lab, FAIR, Reality Labs, infrastructure, ads, ranking, integrity, apps, and GenAI; GenAI comes after the product areas because a third of the GenAI mentions read name the topic, not the former org. The lab's four groups come from stage 1 and overlap. Stage 1 found 238 profiles that name FAIR and not the lab; 215 of them meet the stage-2 definition. The China-educated share is of people who list a bachelor's. A "·" is a cell withheld so that no group of 1 to 9 people can be worked out.

Source: Metix AI Platform, profiles, 2026-09-22.

Queriesmapping.json (population, units) · population.json
POST /v1/people/query · queries/mapping.json (population, units)
{  "population": {    "id": "p2",    "label": "Meta's visible AI staff",    "note": "A current Meta entry (the stage-1 entry: experience.company.name match Meta and experience.is_current eq true), an AI term in the headline or current title, and three filters that apply to everyone except the lab. The lab's 936 people (queries/population.json) are all in P2 by construction, so the lab node is exactly the stage-1 population and no stage-2 count of the lab differs from a stage-1 count by a small group. Everyone else also needs a current Meta entry whose employer size is 10,001+, a current_function outside non_technical_functions, and a current title with none of non_technical_title_words.",    "ai_terms": [      "machine learning",      "ML",      "deep learning",      "computer vision",      "NLP",      "natural language processing",      "LLM",      "LLMs",      "large language models",      "reinforcement learning",      "generative AI",      "GenAI",      "artificial intelligence",      "superintelligence",      "MSL",      "FAIR"    ],    "ai_terms_fields": [      "headline",      "current_title"    ],    "title_only_terms": [      "AI"    ],    "employer_size": "10,001+",    "non_technical_functions": [      "Human Resources",      "Sales",      "Marketing",      "Administrative",      "Finance & Accounting",      "Legal",      "Customer Service",      "Real Estate"    ],    "non_technical_title_words": [      "marketing",      "sales",      "recruiter",      "recruiting",      "sourcer",      "sourcing",      "talent",      "administrative",      "assistant",      "counsel",      "attorney",      "paralegal",      "accountant",      "communications",      "partnerships"    ],    "lab_note": "26 of the 936 list a non-technical function (recruiting and administration, mostly) and 12 more a non-technical title word. They stay in P2 as members of the lab, so the lab row carries them (about 4% of it) while the rest of P2 does not; the function table shows them among other functions.",    "decisions": [      "The draft term list (machine learning, AI, artificial intelligence, research scientist, research engineer, applied scientist, deep learning, LLM, computer vision, NLP, superintelligence, MSL, FAIR, generative AI, GenAI) gave 15,229 current Meta profiles and 14,297 after the function filter.",      "AI alone counts only in the current title. 4,003 draft profiles entered only through AI in the headline, where it is mostly a buzzword (AI-driven testing, cloud and AI platforms); in the first slice read, about half of them held AI roles.",      "Research scientist, research engineer, and applied scientist are not AI terms. Matching is word by word, so research engineer also matched titles such as a software engineer in Reality Labs Research, and plain research titles at Meta also cover people research, survey science, software testing, photonics, materials, and physical modeling. With these titles admitted, reads in California and Washington found 75 to 82% in AI roles; without them, 92%. 2,180 profiles entered only through such a title and are left out; research scientists who name an AI term, FAIR, or the lab stay in.",      "The employer must be the 10,001+ company. Outside the US the name Meta also matches small companies: 6 of 25 profiles in one slice outside the US worked at other firms whose names contain Meta, none of them sized 10,001+. The rule drops 845 draft profiles, most of them Meta work recorded under variant names or through staffing agencies, and a few that are not Meta at all.",      "ML, LLMs, large language models, natural language processing, and reinforcement learning were added; together they bring 547 profiles the draft missed.",      "Title words exclude marketing, sales, recruiting, and other non-technical roles whose current_function is missing or not in the excluded set (administrative, marketing, and sales roles in the reads)."    ],    "audit": "365 profiles read in whole slices, each a narrow filter read in full (one state or country and a range of total_experience_months), never the top of a search. The first slice (49 profiles of the draft) and three candidate slices (53) set the rules above. 150 profiles in eight whole slices fall in the final P2: 138 (92.0%) hold AI roles by their title, headline, and current job description. The last slice, 39 profiles read after the definition was fixed, gave 35 (89.7%) when two borderline creative and advisory roles count as misses. The misses are stale profiles that still list a current Meta job, generic engineers whose headline lists AI words, and a few operations roles. Contract data workers (prompt engineers, annotators, knowledge experts) who list Meta count as AI roles; they appear under data titles. No lab member with a non-technical function falls in the slices read. Records read for the audit are kept private.",    "size_note": "The name Meta on a current entry gives 66,757 profiles in the US and more than 100,000 elsewhere, because outside the US it also matches other companies. With the employer size 10,001+ in the same entry the count is exact on both sides. A company's reported headcount is the comparison; visible profiles also include contractors and profiles not yet updated after someone left."  },  "units": {    "note": "Each person counts at the first unit in this order whose terms appear in the headline or current title, so the rows do not overlap and add up to P2. The lab comes first, so every lab member is in the lab row whatever else they name. Units are what people write, not Meta's org chart.",    "order": [      {        "id": "lab",        "label": "Meta Superintelligence Labs",        "terms": [          "superintelligence",          "MSL"        ]      },      {        "id": "fair",        "label": "FAIR, outside the lab",        "terms": [          "FAIR"        ]      },      {        "id": "reality-labs",        "label": "Reality Labs, AR, VR, and wearables",        "terms": [          "Reality Labs",          "AR",          "VR",          "XR",          "augmented reality",          "virtual reality",          "mixed reality",          "wearables",          "smart glasses",          "Oculus"        ]      },      {        "id": "infrastructure",        "label": "AI infrastructure",        "terms": [          "infra",          "infrastructure"        ]      },      {        "id": "ads",        "label": "Ads and monetization",        "terms": [          "ads",          "monetization",          "advertising"        ]      },      {        "id": "ranking",        "label": "Ranking and recommendations",        "terms": [          "ranking",          "recommendation",          "recommendations",          "recommender",          "recsys"        ]      },      {        "id": "integrity",        "label": "Integrity",        "terms": [          "integrity"        ]      },      {        "id": "apps",        "label": "Instagram, WhatsApp, Messenger, Threads",        "terms": [          "Instagram",          "WhatsApp",          "Messenger",          "Threads"        ]      },      {        "id": "genai",        "label": "GenAI, the former org name",        "terms": [          "GenAI"        ]      }    ],    "none": {      "id": "none",      "label": "Names none of these"    },    "attributes": {      "research": "current_function eq Research",      "engineering": "current_function eq Engineering and Technical",      "manager_up": "current_seniority in Manager, Head, Director, Vice President, President/Vice President, C-Level, Partner, Founder, Owner (the stage-1 set in queries/measures.json)",      "doctorate": "an education entry with degree Doctorate",      "us": "location.country eq United States",      "china_educated": "a Bachelor entry at an institution on the mainland-China list (china_educated below)",      "bachelor": "any Bachelor entry, the base of the China-educated share"    },    "audit": [      "GenAI: a whole slice of 24 profiles outside the lab: 16 name Meta's former GenAI org, and 8 use the word for the topic, 5 of them inside ads, integrity, Instagram, or AR and VR work. GenAI therefore comes after the product areas, so that topic uses that name an area count there.",      "Generative AI without GenAI: a whole slice of 23: 5 read as the org, all of them job titles, and 18 as the topic. The phrase is not a unit term.",      "AR, VR, and wearables without Reality Labs: a whole slice of 16: 12 describe work on Reality Labs products (Quest, smart glasses, wearables), 4 do not (ads, sales, creator, and stale profiles). Reality Labs itself is the org name.",      "Infra or infrastructure outside the lab: a whole slice of 28: 26 work on AI infrastructure, 6 of them ads ML infrastructure, which counts here because infrastructure comes before ads; 2 work on network or general infrastructure.",      "FAIR outside the lab: all 7 who name it in the eight P2 slices are FAIR research staff. 215 of the 238 current Meta profiles that name FAIR and no lab term (stage 1) are in P2.",      "Ads, ranking, integrity, and app names were not read in slices of their own; in the P2 slices they name the current team's product area (ads ranking, Reels ranking, integrity, Instagram)."    ]  }}
POST /v1/people/query · queries/population.json
{  "note": "Who is counted. A person counts when one job entry is at Meta and current (experience.company.name match current_employer and experience.is_current eq true, in one has_experience entry) and the profile's headline or current_title names the lab with one of lab_terms. The Platform matches text word by word, in any order, and ignores case, so MSL also matches msl and superintelligence also matches SuperIntelligence. Counts are profiles visible through the Metix AI Platform, never the lab's headcount; stale profiles count as current, so a count is not a guaranteed floor either.",  "current_employer": "Meta",  "current_employer_note": "match Meta on the current entry. The name also matches entries written as Facebook: a current entry named Facebook counts, and a Facebook condition returned all 790 profiles of the narrow definition, so the two names are one employer on the Platform and a Facebook condition is no separate test. company.name eq Meta returned the same 790 profiles as match Meta, and every one of the 63 profiles read in the audit has a current entry at Meta itself.",  "lab_terms": [    "superintelligence",    "MSL"  ],  "lab_fields": [    "headline",    "current_title"  ],  "lab_terms_note": "superintelligence alone (the narrow definition) gives 790 profiles. MSL, the lab's short name, adds 146 current Meta profiles that never write superintelligence; 13 of 14 read in a whole slice use it for the lab, and the rest recruit for it. FAIR, one of the lab's groups since 2025, is not a lab term: people who name only FAIR do not name the lab, and teams.json counts them as context (fair_only).",  "narrow_terms": [    "superintelligence"  ],  "history_names": [    "Meta",    "Facebook",    "Instagram",    "WhatsApp",    "Oculus"  ],  "history_names_note": "Names that count as an earlier job at Meta, matched word by word in one has_experience entry. The name Meta does not match an entry listed under Instagram; adding the other names moves fewer than 10 people.",  "cut": "2025-06-01",  "cut_note": "Meta announced Meta Superintelligence Labs at the end of June 2025. A person with a job entry at a history name that started before the cut had a Meta job before the lab. experience.start_date is a month (YYYY-MM) or a year; the Platform compares it as its first day, so a year alone reads as January. A newcomer whose Meta entry says only 2025 therefore counts as at Meta before the cut, which raises the share at Meta when the lab formed, and an earlier entry whose end date says only 2025 counts as ending before May 2025, which can move someone who was at Meta into returned.",  "stay_end": "2025-05-01",  "cohorts_note": "Four groups that do not overlap and add up to the whole. Stayed: a non-intern Meta entry that started before the cut and is current or ended in May 2025 or later, so the person was at Meta when the lab formed. Returned: a non-intern Meta entry before the cut, but none that lasted into May 2025, so they had left and came back. Former interns: the only Meta entries before the cut are internships. New to Meta: no Meta entry that started before the cut, including fewer than 10 whose current Meta entry has no start date. Returned, former interns, and new to Meta together are the outside hires, everyone who was not at Meta when the lab formed.",  "intern_rule": "An entry is an internship when experience.seniority eq Intern or experience.title matches intern. Both are tested inside the same has_experience entry as the employer and the date.",  "labeling_terms": [    "labeling",    "labelling",    "annotation",    "annotator",    "red teamer",    "red teaming"  ],  "labeling_note": "Titles and headlines that name data labeling, annotation, or red teaming. The audit found such roles among recent hires (fewer than 10 in the slice read); vendors often staff them, but the profile lists Meta.",  "freshness_note": "Profiles are a periodically refreshed snapshot. Across all Meta job entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so start dates stop in April 2026 and the most recent months are probably undercounted. April 2026 is the last month shown, and people who joined since are missing.",  "audit": "63 profiles read in whole slices, never from the top of a search, because Search ranks by match quality. The slices covered people with no Meta-named entry before the cut, people with a Meta entry before the cut, people who write MSL but not superintelligence, and the profiles whose Meta entry has no start date (fewer than 10). 62 of 63 describe themselves as working in the lab, and the rest recruit for it. All 63 have a current Meta entry, almost all of them named Meta. All 31 read with a Meta entry before the cut had a real Meta job there, not only an internship. Among people without a Meta-named entry before the cut, an earlier Meta job turned up listed under Instagram, which is why Instagram is a history name, and fewer than 10 of them hold data labeling or red teaming roles. Separate reads of sub-team matches found fewer than 10 profiles that say they have left the lab while the Meta entry is still current, so a few members are out of date. Records read for the audit are kept private.",  "comparison": {    "id": "microsoft",    "label": "Microsoft",    "current_employer": "Microsoft",    "lab_terms": [      "superintelligence"    ],    "history_names": [      "Microsoft"    ],    "cut": "2025-11-01",    "stay_end": "2025-10-01",    "edges": [      "2012-01-01",      "2016-01-01",      "2020-01-01",      "2023-01-01",      "2025-01-01"    ],    "note": "Current Microsoft staff whose headline or current_title names superintelligence. Microsoft AI announced its superintelligence team in November 2025, so the cut for this group is 2025-11-01. Microsoft also matches Microsoft AI and Microsoft Research; subsidiaries with their own names (LinkedIn, GitHub) are not history names. A whole slice of 15 all named the Microsoft AI superintelligence team; fewer than 10 of them are recruiters or sourcers for it."  }}

11

Profiles at Meta outnumber its reported headcount 1.7 to 1; these count profiles, not staff

The bars share one scale, set by the largest count, the current Meta profiles

Visible profiles 128,970; reported headcount 75,472; naming an AI role 8,257; naming the lab 936

  1. Current Meta profiles on the Platformobserved128,970
  2. Meta's reported headcount, June 30, 2026public, SEC filing75,472
  3. Profiles that name an AI roleobserved8,257
  4. Of them, naming the labobserved936

What it shows

Meta reported 75,472 employees on June 30, 2026, and the Platform shows 128,970 current Meta profiles, 1.7 for each employee. Profiles of people who left without updating, and of contractors who list Meta, count as current, so a count here is of profiles, and the 8,257 is neither a floor nor a ceiling on Meta's AI staff.

Method and limits

The profile count is current entries at Meta, the 10,001+ company, in and outside the US. The headcount is Meta's own figure (SEC filing, July 29, 2026), which includes about 8,000 employees affected by the May 2026 reduction; profile dates stop in April 2026, so the data does not show that reduction.

Source: Metix AI Platform, profiles and jobs, 2026-09-22.

Querymapping.json (population)
POST /v1/people/query · queries/mapping.json (population)
{  "population": {    "id": "p2",    "label": "Meta's visible AI staff",    "note": "A current Meta entry (the stage-1 entry: experience.company.name match Meta and experience.is_current eq true), an AI term in the headline or current title, and three filters that apply to everyone except the lab. The lab's 936 people (queries/population.json) are all in P2 by construction, so the lab node is exactly the stage-1 population and no stage-2 count of the lab differs from a stage-1 count by a small group. Everyone else also needs a current Meta entry whose employer size is 10,001+, a current_function outside non_technical_functions, and a current title with none of non_technical_title_words.",    "ai_terms": [      "machine learning",      "ML",      "deep learning",      "computer vision",      "NLP",      "natural language processing",      "LLM",      "LLMs",      "large language models",      "reinforcement learning",      "generative AI",      "GenAI",      "artificial intelligence",      "superintelligence",      "MSL",      "FAIR"    ],    "ai_terms_fields": [      "headline",      "current_title"    ],    "title_only_terms": [      "AI"    ],    "employer_size": "10,001+",    "non_technical_functions": [      "Human Resources",      "Sales",      "Marketing",      "Administrative",      "Finance & Accounting",      "Legal",      "Customer Service",      "Real Estate"    ],    "non_technical_title_words": [      "marketing",      "sales",      "recruiter",      "recruiting",      "sourcer",      "sourcing",      "talent",      "administrative",      "assistant",      "counsel",      "attorney",      "paralegal",      "accountant",      "communications",      "partnerships"    ],    "lab_note": "26 of the 936 list a non-technical function (recruiting and administration, mostly) and 12 more a non-technical title word. They stay in P2 as members of the lab, so the lab row carries them (about 4% of it) while the rest of P2 does not; the function table shows them among other functions.",    "decisions": [      "The draft term list (machine learning, AI, artificial intelligence, research scientist, research engineer, applied scientist, deep learning, LLM, computer vision, NLP, superintelligence, MSL, FAIR, generative AI, GenAI) gave 15,229 current Meta profiles and 14,297 after the function filter.",      "AI alone counts only in the current title. 4,003 draft profiles entered only through AI in the headline, where it is mostly a buzzword (AI-driven testing, cloud and AI platforms); in the first slice read, about half of them held AI roles.",      "Research scientist, research engineer, and applied scientist are not AI terms. Matching is word by word, so research engineer also matched titles such as a software engineer in Reality Labs Research, and plain research titles at Meta also cover people research, survey science, software testing, photonics, materials, and physical modeling. With these titles admitted, reads in California and Washington found 75 to 82% in AI roles; without them, 92%. 2,180 profiles entered only through such a title and are left out; research scientists who name an AI term, FAIR, or the lab stay in.",      "The employer must be the 10,001+ company. Outside the US the name Meta also matches small companies: 6 of 25 profiles in one slice outside the US worked at other firms whose names contain Meta, none of them sized 10,001+. The rule drops 845 draft profiles, most of them Meta work recorded under variant names or through staffing agencies, and a few that are not Meta at all.",      "ML, LLMs, large language models, natural language processing, and reinforcement learning were added; together they bring 547 profiles the draft missed.",      "Title words exclude marketing, sales, recruiting, and other non-technical roles whose current_function is missing or not in the excluded set (administrative, marketing, and sales roles in the reads)."    ],    "audit": "365 profiles read in whole slices, each a narrow filter read in full (one state or country and a range of total_experience_months), never the top of a search. The first slice (49 profiles of the draft) and three candidate slices (53) set the rules above. 150 profiles in eight whole slices fall in the final P2: 138 (92.0%) hold AI roles by their title, headline, and current job description. The last slice, 39 profiles read after the definition was fixed, gave 35 (89.7%) when two borderline creative and advisory roles count as misses. The misses are stale profiles that still list a current Meta job, generic engineers whose headline lists AI words, and a few operations roles. Contract data workers (prompt engineers, annotators, knowledge experts) who list Meta count as AI roles; they appear under data titles. No lab member with a non-technical function falls in the slices read. Records read for the audit are kept private.",    "size_note": "The name Meta on a current entry gives 66,757 profiles in the US and more than 100,000 elsewhere, because outside the US it also matches other companies. With the employer size 10,001+ in the same entry the count is exact on both sides. A company's reported headcount is the comparison; visible profiles also include contractors and profiles not yet updated after someone left."  }}

The work

Part 5, Figures 12 to 14

What the lab and the rest of Meta's AI staff do, at what level, under which titles, and in which technical directions, against what Meta's postings ask for.

12

35.1% of the lab work in research, against 17.8% of the rest of Meta's visible AI staff

Current function and level as a share of each group, on a 0 to 60% scale: the lab (936), the rest of Meta's visible AI staff (7,321), and Meta's AI postings (218), placed by title family, with no level

Research: 35.1% / 17.8% / 23.9%; Engineering: 37.3% / 58.0% / 49.1%; Product, design, and program: 10.6% / 8.0% / 6.4%; Data: 3.2% / 6.3% / 4.1%; Other, or none stated: 13.8% / 9.8% / 16.5%; Specialist or intern: 38.7% / 50.0%; Senior: 16.5% / 13.5%; Manager: 12.3% / 10.4%; Head, director, and above: 5.3% / 3.3%; None stated: 27.2% / 22.8%

What it shows

The lab is 35.1% research and 37.3% engineering; the rest of Meta's AI staff are 17.8% and 58.0%. Among Meta's AI postings, 49.1% are engineering titles and 23.9% research titles.

By level, 17.6% of the lab are manager or above, against 13.7% of the rest; heads, directors, and above are 5.3% against 3.3%.

Method and limits

Function is the Platform's fixed value for the current job; where it is not Research, data titles are taken out first. About 4% of the lab list a non-technical function or title and stay in the lab so it matches stage 1; the lab's other and none-stated cells are merged so that no stage-1 cell can be subtracted down to a small group. Postings are placed by title: research scientist and research engineer as research, machine learning and software engineer as engineering. The comparison is uneven by construction: the lab counts everyone who names it, while the rest counts only profiles that name an AI term, which leaves out about 2,200 research-titled profiles with no AI term. If many of those hold AI roles, the rest's research share is higher and the gap smaller.

Level is also the Platform's fixed value: a visible hierarchy inferred from titles, not a verified reporting line. 27.2% of the lab and 22.8% of the rest state none. The lab has fewer than 10 interns and fewer than 10 people above director, so interns join specialists and every level from head up is one group.

Source: Metix AI Platform, profiles and jobs, 2026-09-22.

Querymapping.json (population, functions, levels, postings)
POST /v1/people/query · POST /v1/jobs/query · queries/mapping.json (population, functions, levels, postings)
{  "population": {    "id": "p2",    "label": "Meta's visible AI staff",    "note": "A current Meta entry (the stage-1 entry: experience.company.name match Meta and experience.is_current eq true), an AI term in the headline or current title, and three filters that apply to everyone except the lab. The lab's 936 people (queries/population.json) are all in P2 by construction, so the lab node is exactly the stage-1 population and no stage-2 count of the lab differs from a stage-1 count by a small group. Everyone else also needs a current Meta entry whose employer size is 10,001+, a current_function outside non_technical_functions, and a current title with none of non_technical_title_words.",    "ai_terms": [      "machine learning",      "ML",      "deep learning",      "computer vision",      "NLP",      "natural language processing",      "LLM",      "LLMs",      "large language models",      "reinforcement learning",      "generative AI",      "GenAI",      "artificial intelligence",      "superintelligence",      "MSL",      "FAIR"    ],    "ai_terms_fields": [      "headline",      "current_title"    ],    "title_only_terms": [      "AI"    ],    "employer_size": "10,001+",    "non_technical_functions": [      "Human Resources",      "Sales",      "Marketing",      "Administrative",      "Finance & Accounting",      "Legal",      "Customer Service",      "Real Estate"    ],    "non_technical_title_words": [      "marketing",      "sales",      "recruiter",      "recruiting",      "sourcer",      "sourcing",      "talent",      "administrative",      "assistant",      "counsel",      "attorney",      "paralegal",      "accountant",      "communications",      "partnerships"    ],    "lab_note": "26 of the 936 list a non-technical function (recruiting and administration, mostly) and 12 more a non-technical title word. They stay in P2 as members of the lab, so the lab row carries them (about 4% of it) while the rest of P2 does not; the function table shows them among other functions.",    "decisions": [      "The draft term list (machine learning, AI, artificial intelligence, research scientist, research engineer, applied scientist, deep learning, LLM, computer vision, NLP, superintelligence, MSL, FAIR, generative AI, GenAI) gave 15,229 current Meta profiles and 14,297 after the function filter.",      "AI alone counts only in the current title. 4,003 draft profiles entered only through AI in the headline, where it is mostly a buzzword (AI-driven testing, cloud and AI platforms); in the first slice read, about half of them held AI roles.",      "Research scientist, research engineer, and applied scientist are not AI terms. Matching is word by word, so research engineer also matched titles such as a software engineer in Reality Labs Research, and plain research titles at Meta also cover people research, survey science, software testing, photonics, materials, and physical modeling. With these titles admitted, reads in California and Washington found 75 to 82% in AI roles; without them, 92%. 2,180 profiles entered only through such a title and are left out; research scientists who name an AI term, FAIR, or the lab stay in.",      "The employer must be the 10,001+ company. Outside the US the name Meta also matches small companies: 6 of 25 profiles in one slice outside the US worked at other firms whose names contain Meta, none of them sized 10,001+. The rule drops 845 draft profiles, most of them Meta work recorded under variant names or through staffing agencies, and a few that are not Meta at all.",      "ML, LLMs, large language models, natural language processing, and reinforcement learning were added; together they bring 547 profiles the draft missed.",      "Title words exclude marketing, sales, recruiting, and other non-technical roles whose current_function is missing or not in the excluded set (administrative, marketing, and sales roles in the reads)."    ],    "audit": "365 profiles read in whole slices, each a narrow filter read in full (one state or country and a range of total_experience_months), never the top of a search. The first slice (49 profiles of the draft) and three candidate slices (53) set the rules above. 150 profiles in eight whole slices fall in the final P2: 138 (92.0%) hold AI roles by their title, headline, and current job description. The last slice, 39 profiles read after the definition was fixed, gave 35 (89.7%) when two borderline creative and advisory roles count as misses. The misses are stale profiles that still list a current Meta job, generic engineers whose headline lists AI words, and a few operations roles. Contract data workers (prompt engineers, annotators, knowledge experts) who list Meta count as AI roles; they appear under data titles. No lab member with a non-technical function falls in the slices read. Records read for the audit are kept private.",    "size_note": "The name Meta on a current entry gives 66,757 profiles in the US and more than 100,000 elsewhere, because outside the US it also matches other companies. With the employer size 10,001+ in the same entry the count is exact on both sides. A company's reported headcount is the comparison; visible profiles also include contractors and profiles not yet updated after someone left."  },  "functions": {    "note": "Five functions and none, exclusive, in this order: research (current_function Research), data (a data title and not Research), engineering (Engineering and Technical, no data title), product (Product, Design, or Project Management, no data title), other stated, none stated. The lab and the rest of P2 each add up to their total.",    "data_title_words": [      "data",      "annotation",      "annotations",      "annotator",      "labeling",      "labelling",      "labeler",      "prompt",      "knowledge expert",      "rater",      "evaluator"    ],    "product_functions": [      "Product",      "Design",      "Project Management"    ],    "audit": "Data titles in the reads: data engineers and data scientists, annotators and data labeling analysts, prompt engineers, knowledge experts, and AI data quality reviewers, many of them contract workers. Several annotators carry the Research function, which is why research comes first and keeps them: the data row is a lower bound on data work.",    "lab_rule": "The lab's functions are also published in stage 1 (data/roles.json). When a difference between a stage-1 cell and a stage-2 cell would be 1 to 9 people, the lab row merges other and none into one cell; if a difference is still 1 to 9, the run stops without writing."  },  "levels": {    "note": "current_seniority is counted in the groups below (Vice President and above is manager_up minus Manager, Head, and Director) and published in five: specialist-or-intern, senior, manager, head-and-above (Head, Director, and every level above), and none stated. The lab holds fewer than 10 interns and fewer than 10 above Director, and with the lab's manager_up published a reader could work out both, so Intern joins Specialist (as in stage 1) and the levels from Head up form one group. A visible hierarchy, not a reporting line.",    "groups": [      {        "id": "intern",        "values": [          "Intern"        ]      },      {        "id": "specialist",        "values": [          "Specialist"        ]      },      {        "id": "senior",        "values": [          "Senior"        ]      },      {        "id": "manager",        "values": [          "Manager"        ]      },      {        "id": "head-director",        "values": [          "Head",          "Director"        ]      },      {        "id": "vp-up",        "values": [          "Vice President",          "President/Vice President",          "C-Level",          "Partner",          "Founder",          "Owner"        ]      }    ],    "individual_contributor": [      "Intern",      "Specialist",      "Senior"    ],    "published": [      "specialist-or-intern",      "senior",      "manager",      "head-and-above",      "none"    ]  },  "postings": {    "note": "Meta postings (company.name match Meta; the jobs index has no employer size, so a few postings at other companies named Meta may be included) posted in the last 180 days (posted_date gte now-180d) whose title contains a P2 AI term or AI, and none of the non-technical title words. The index holds open postings only, and every one read was posted within three weeks of the snapshot, so this is the demand open on the day. A role posted in several cities counts once per city. Counts of postings carry no minimum.",    "posted_from": "now-180d",    "families": [      {        "id": "research-scientist",        "label": "Research scientist",        "title": [          "research scientist"        ]      },      {        "id": "research-engineer",        "label": "Research engineer",        "title": [          "research engineer"        ]      },      {        "id": "ml-software-engineer",        "label": "Machine learning or software engineer",        "title": [          "software engineer",          "machine learning engineer",          "ML engineer",          "AI engineer"        ]      },      {        "id": "data",        "label": "Data",        "title": [          "data",          "annotation",          "annotator",          "labeling",          "prompt",          "evaluator"        ]      },      {        "id": "product-program",        "label": "Product, program, and design",        "title": [          "product manager",          "product management",          "program manager",          "designer"        ]      }    ],    "families_note": "Exclusive, first match in this order; the rest are other (hardware, production, security, network, and leadership titles).",    "audit": "91 read (see directions): 73 pass the title rule. Plain research scientist postings include demography and survey science, server demand forecasting, and infrastructure reliability, and AI in a title also names marketing and business development roles; the rule drops those. The 73 left are AI roles, a few of them borderline (security for AI, analytics for AI ranking, AI hardware bring-up). 90 of 91 state a range in USD."  }}

13

24.2% of the rest are titled machine learning engineer, against 3.8% in the lab

People whose current title contains the words, as a share of each group, on a 0 to 30% scale. A title can count in several rows

Machine learning engineer: 3.8% / 24.2%; Software engineer: 24.4% / 20.5%; Research scientist: 25.3% / 10.4%; Engineering manager: 4.3% / 4.3%; Director: 5.6% / 2.3%; Product manager: 4.0% / 2.4%; Research engineer: 5.6% / 2.1%; Data scientist: 2.6% / 2.4%; Program manager: <1% / 2.3%; Data engineer: 1.5% / 1.1%

What it shows

In the lab, the commonest titles are research scientist (25.3%) and software engineer (24.4%); among the rest, 10.4% are research scientists.

Method and limits

A title counts when it contains all the words of a candidate, so rows overlap (Software Engineer, Machine Learning counts in two rows). The candidates are the titles that recur in the audit reads; the ten with the most profiles are shown. The rest leaves out research titles with no AI term (figure 12), which raises its machine learning engineer share.

Source: Metix AI Platform, profiles, 2026-09-22.

Querymapping.json (population, titles)
POST /v1/people/query · queries/mapping.json (population, titles)
{  "population": {    "id": "p2",    "label": "Meta's visible AI staff",    "note": "A current Meta entry (the stage-1 entry: experience.company.name match Meta and experience.is_current eq true), an AI term in the headline or current title, and three filters that apply to everyone except the lab. The lab's 936 people (queries/population.json) are all in P2 by construction, so the lab node is exactly the stage-1 population and no stage-2 count of the lab differs from a stage-1 count by a small group. Everyone else also needs a current Meta entry whose employer size is 10,001+, a current_function outside non_technical_functions, and a current title with none of non_technical_title_words.",    "ai_terms": [      "machine learning",      "ML",      "deep learning",      "computer vision",      "NLP",      "natural language processing",      "LLM",      "LLMs",      "large language models",      "reinforcement learning",      "generative AI",      "GenAI",      "artificial intelligence",      "superintelligence",      "MSL",      "FAIR"    ],    "ai_terms_fields": [      "headline",      "current_title"    ],    "title_only_terms": [      "AI"    ],    "employer_size": "10,001+",    "non_technical_functions": [      "Human Resources",      "Sales",      "Marketing",      "Administrative",      "Finance & Accounting",      "Legal",      "Customer Service",      "Real Estate"    ],    "non_technical_title_words": [      "marketing",      "sales",      "recruiter",      "recruiting",      "sourcer",      "sourcing",      "talent",      "administrative",      "assistant",      "counsel",      "attorney",      "paralegal",      "accountant",      "communications",      "partnerships"    ],    "lab_note": "26 of the 936 list a non-technical function (recruiting and administration, mostly) and 12 more a non-technical title word. They stay in P2 as members of the lab, so the lab row carries them (about 4% of it) while the rest of P2 does not; the function table shows them among other functions.",    "decisions": [      "The draft term list (machine learning, AI, artificial intelligence, research scientist, research engineer, applied scientist, deep learning, LLM, computer vision, NLP, superintelligence, MSL, FAIR, generative AI, GenAI) gave 15,229 current Meta profiles and 14,297 after the function filter.",      "AI alone counts only in the current title. 4,003 draft profiles entered only through AI in the headline, where it is mostly a buzzword (AI-driven testing, cloud and AI platforms); in the first slice read, about half of them held AI roles.",      "Research scientist, research engineer, and applied scientist are not AI terms. Matching is word by word, so research engineer also matched titles such as a software engineer in Reality Labs Research, and plain research titles at Meta also cover people research, survey science, software testing, photonics, materials, and physical modeling. With these titles admitted, reads in California and Washington found 75 to 82% in AI roles; without them, 92%. 2,180 profiles entered only through such a title and are left out; research scientists who name an AI term, FAIR, or the lab stay in.",      "The employer must be the 10,001+ company. Outside the US the name Meta also matches small companies: 6 of 25 profiles in one slice outside the US worked at other firms whose names contain Meta, none of them sized 10,001+. The rule drops 845 draft profiles, most of them Meta work recorded under variant names or through staffing agencies, and a few that are not Meta at all.",      "ML, LLMs, large language models, natural language processing, and reinforcement learning were added; together they bring 547 profiles the draft missed.",      "Title words exclude marketing, sales, recruiting, and other non-technical roles whose current_function is missing or not in the excluded set (administrative, marketing, and sales roles in the reads)."    ],    "audit": "365 profiles read in whole slices, each a narrow filter read in full (one state or country and a range of total_experience_months), never the top of a search. The first slice (49 profiles of the draft) and three candidate slices (53) set the rules above. 150 profiles in eight whole slices fall in the final P2: 138 (92.0%) hold AI roles by their title, headline, and current job description. The last slice, 39 profiles read after the definition was fixed, gave 35 (89.7%) when two borderline creative and advisory roles count as misses. The misses are stale profiles that still list a current Meta job, generic engineers whose headline lists AI words, and a few operations roles. Contract data workers (prompt engineers, annotators, knowledge experts) who list Meta count as AI roles; they appear under data titles. No lab member with a non-technical function falls in the slices read. Records read for the audit are kept private.",    "size_note": "The name Meta on a current entry gives 66,757 profiles in the US and more than 100,000 elsewhere, because outside the US it also matches other companies. With the employer size 10,001+ in the same entry the count is exact on both sides. A company's reported headcount is the comparison; visible profiles also include contractors and profiles not yet updated after someone left."  },  "titles": {    "note": "Current titles that contain all the words of a candidate, in any order, so rows overlap: Software Engineer, Machine Learning counts under both software engineer and machine learning engineer. Candidates are the titles that recur in the audit reads; the ten with the most P2 profiles are published.",    "candidates": [      "machine learning engineer",      "software engineer",      "research scientist",      "research engineer",      "engineering manager",      "product manager",      "program manager",      "data scientist",      "data engineer",      "prompt engineer",      "production engineer",      "director"    ],    "publish": 10  }}

14

Postings name ranking, LLMs, and infrastructure most; the lab's staff lean most to LLMs

Left: staff who name the direction, as a share of each group, on a 0 to 30% scale. Right: postings that name it, as a share of Meta's 218 AI postings, on a 0 to 50% scale. Sorted by postings

Ranking and recommendations: 8.2% / 17.9%; postings 46.3%. Foundation models and LLMs: 28.7% / 18.5%; postings 35.8%. AI infrastructure: 13.1% / 15.5%; postings 34.9%. Agents: 5.7% / 5.0%; postings 17.4%. Post-training and alignment: 8.8% / 6.4%; postings 14.7%. Multimodal and perception: 15.9% / 12.8%; postings 12.4%. AR, VR, and devices: 2.7% / 7.3%; postings 6.4%. Safety and evaluation: 4.2% / 3.6%; postings 4.1%

What it shows

Postings name ranking and recommendations most (46.3%), then LLMs (35.8%) and AI infrastructure (34.9%). Among staff, the lab leans to foundation models and LLMs (28.7% against 18.5%), and the rest to ranking (17.9% against 8.2%).

Method and limits

A profile or a posting can name several directions. Staff are matched on headline, title, and skills; postings on title and description. Descriptions run longer than headlines, so compare the order on each side, not the levels. Postings are Meta's 218 AI postings still open on the snapshot day, posted in the last 180 days, and a role posted in several cities counts once per city; the jobs index has no employer size, so a few postings at other companies named Meta may be included. Every description carries a standard paragraph on AR and VR, so descriptions match only on distinctive single words.

Source: Metix AI Platform, profiles and jobs, 2026-09-22.

Querymapping.json (population, directions, postings)
POST /v1/people/query · POST /v1/jobs/query · queries/mapping.json (population, directions, postings)
{  "population": {    "id": "p2",    "label": "Meta's visible AI staff",    "note": "A current Meta entry (the stage-1 entry: experience.company.name match Meta and experience.is_current eq true), an AI term in the headline or current title, and three filters that apply to everyone except the lab. The lab's 936 people (queries/population.json) are all in P2 by construction, so the lab node is exactly the stage-1 population and no stage-2 count of the lab differs from a stage-1 count by a small group. Everyone else also needs a current Meta entry whose employer size is 10,001+, a current_function outside non_technical_functions, and a current title with none of non_technical_title_words.",    "ai_terms": [      "machine learning",      "ML",      "deep learning",      "computer vision",      "NLP",      "natural language processing",      "LLM",      "LLMs",      "large language models",      "reinforcement learning",      "generative AI",      "GenAI",      "artificial intelligence",      "superintelligence",      "MSL",      "FAIR"    ],    "ai_terms_fields": [      "headline",      "current_title"    ],    "title_only_terms": [      "AI"    ],    "employer_size": "10,001+",    "non_technical_functions": [      "Human Resources",      "Sales",      "Marketing",      "Administrative",      "Finance & Accounting",      "Legal",      "Customer Service",      "Real Estate"    ],    "non_technical_title_words": [      "marketing",      "sales",      "recruiter",      "recruiting",      "sourcer",      "sourcing",      "talent",      "administrative",      "assistant",      "counsel",      "attorney",      "paralegal",      "accountant",      "communications",      "partnerships"    ],    "lab_note": "26 of the 936 list a non-technical function (recruiting and administration, mostly) and 12 more a non-technical title word. They stay in P2 as members of the lab, so the lab row carries them (about 4% of it) while the rest of P2 does not; the function table shows them among other functions.",    "decisions": [      "The draft term list (machine learning, AI, artificial intelligence, research scientist, research engineer, applied scientist, deep learning, LLM, computer vision, NLP, superintelligence, MSL, FAIR, generative AI, GenAI) gave 15,229 current Meta profiles and 14,297 after the function filter.",      "AI alone counts only in the current title. 4,003 draft profiles entered only through AI in the headline, where it is mostly a buzzword (AI-driven testing, cloud and AI platforms); in the first slice read, about half of them held AI roles.",      "Research scientist, research engineer, and applied scientist are not AI terms. Matching is word by word, so research engineer also matched titles such as a software engineer in Reality Labs Research, and plain research titles at Meta also cover people research, survey science, software testing, photonics, materials, and physical modeling. With these titles admitted, reads in California and Washington found 75 to 82% in AI roles; without them, 92%. 2,180 profiles entered only through such a title and are left out; research scientists who name an AI term, FAIR, or the lab stay in.",      "The employer must be the 10,001+ company. Outside the US the name Meta also matches small companies: 6 of 25 profiles in one slice outside the US worked at other firms whose names contain Meta, none of them sized 10,001+. The rule drops 845 draft profiles, most of them Meta work recorded under variant names or through staffing agencies, and a few that are not Meta at all.",      "ML, LLMs, large language models, natural language processing, and reinforcement learning were added; together they bring 547 profiles the draft missed.",      "Title words exclude marketing, sales, recruiting, and other non-technical roles whose current_function is missing or not in the excluded set (administrative, marketing, and sales roles in the reads)."    ],    "audit": "365 profiles read in whole slices, each a narrow filter read in full (one state or country and a range of total_experience_months), never the top of a search. The first slice (49 profiles of the draft) and three candidate slices (53) set the rules above. 150 profiles in eight whole slices fall in the final P2: 138 (92.0%) hold AI roles by their title, headline, and current job description. The last slice, 39 profiles read after the definition was fixed, gave 35 (89.7%) when two borderline creative and advisory roles count as misses. The misses are stale profiles that still list a current Meta job, generic engineers whose headline lists AI words, and a few operations roles. Contract data workers (prompt engineers, annotators, knowledge experts) who list Meta count as AI roles; they appear under data titles. No lab member with a non-technical function falls in the slices read. Records read for the audit are kept private.",    "size_note": "The name Meta on a current entry gives 66,757 profiles in the US and more than 100,000 elsewhere, because outside the US it also matches other companies. With the employer size 10,001+ in the same entry the count is exact on both sides. A company's reported headcount is the comparison; visible profiles also include contractors and profiles not yet updated after someone left."  },  "directions": {    "note": "Multi-label. Stock: P2 profiles whose headline or current title contains a phrase, or whose skills contain a single word, of the direction. Flow: Meta AI postings (postings below) whose title contains a phrase or whose description contains a single word. Long fields (skills lists, descriptions) take single distinctive words only, because a phrase matches its words anywhere in the text.",    "list": [      {        "id": "foundation-models",        "label": "Foundation models and LLMs",        "phrase": [          "LLM",          "LLMs",          "large language model",          "large language models",          "foundation model",          "foundation models",          "language model",          "language models",          "pre-training",          "pretraining",          "Llama"        ],        "word": [          "LLM",          "LLMs",          "pretraining",          "Llama"        ]      },      {        "id": "post-training",        "label": "Post-training and alignment",        "phrase": [          "post-training",          "post training",          "RLHF",          "reinforcement learning",          "fine-tuning",          "alignment"        ],        "word": [          "RLHF",          "RLVR",          "reinforcement"        ]      },      {        "id": "multimodal-perception",        "label": "Multimodal and perception",        "phrase": [          "computer vision",          "vision",          "multimodal",          "speech",          "video",          "image",          "perception",          "3D",          "VLM"        ],        "word": [          "multimodal",          "VLM",          "VLMs",          "speech"        ]      },      {        "id": "ai-infrastructure",        "label": "AI infrastructure",        "phrase": [          "infra",          "infrastructure",          "GPU",          "GPUs",          "CUDA",          "inference",          "compiler",          "compilers",          "accelerator",          "accelerators",          "kernels",          "distributed training",          "systems ML",          "SystemML",          "HPC",          "MTIA",          "PyTorch"        ],        "word": [          "GPU",          "GPUs",          "CUDA",          "compiler",          "compilers",          "accelerator",          "accelerators",          "kernels",          "MTIA",          "HPC"        ]      },      {        "id": "agents",        "label": "Agents",        "phrase": [          "agent",          "agents",          "agentic"        ],        "word": [          "agentic",          "agents"        ]      },      {        "id": "ranking-recommendations",        "label": "Ranking and recommendations",        "phrase": [          "ranking",          "recommendation",          "recommendations",          "recommender",          "recsys"        ],        "word": [          "ranking",          "recommendation",          "recommender",          "recsys"        ]      },      {        "id": "ar-vr-devices",        "label": "AR, VR, and devices",        "phrase": [          "AR",          "VR",          "XR",          "augmented reality",          "virtual reality",          "mixed reality",          "wearables",          "smart glasses",          "Reality Labs",          "Oculus"        ],        "word": [          "wearables",          "XR",          "glasses"        ]      },      {        "id": "safety-evaluation",        "label": "Safety and evaluation",        "phrase": [          "safety",          "evaluation",          "evals",          "red teaming",          "responsible AI"        ],        "word": [          "evals",          "teaming"        ]      }    ],    "audit": "91 postings read in two whole slices (every Meta AI-titled posting of the last 180 days in New York, 30, and in Menlo Park, 61). Every description carries Meta's standard paragraph on augmented and virtual reality, and 59 of 91 mention responsible AI, so those phrases would match nearly every posting. Alignment appears mostly as stakeholder alignment, post-training also matches post-silicon in accelerator postings that mention training, inference also means causal inference, and computer vision matches any description with Computer Science and vision. Hence single distinctive words for descriptions; with them the matches read as the direction, except agents, which one security posting uses for its own tooling. In skills, safety mostly meant workplace safety and perception included brand perception, so neither is a skill word."  },  "postings": {    "note": "Meta postings (company.name match Meta; the jobs index has no employer size, so a few postings at other companies named Meta may be included) posted in the last 180 days (posted_date gte now-180d) whose title contains a P2 AI term or AI, and none of the non-technical title words. The index holds open postings only, and every one read was posted within three weeks of the snapshot, so this is the demand open on the day. A role posted in several cities counts once per city. Counts of postings carry no minimum.",    "posted_from": "now-180d",    "families": [      {        "id": "research-scientist",        "label": "Research scientist",        "title": [          "research scientist"        ]      },      {        "id": "research-engineer",        "label": "Research engineer",        "title": [          "research engineer"        ]      },      {        "id": "ml-software-engineer",        "label": "Machine learning or software engineer",        "title": [          "software engineer",          "machine learning engineer",          "ML engineer",          "AI engineer"        ]      },      {        "id": "data",        "label": "Data",        "title": [          "data",          "annotation",          "annotator",          "labeling",          "prompt",          "evaluator"        ]      },      {        "id": "product-program",        "label": "Product, program, and design",        "title": [          "product manager",          "product management",          "program manager",          "designer"        ]      }    ],    "families_note": "Exclusive, first match in this order; the rest are other (hardware, production, security, network, and leadership titles).",    "audit": "91 read (see directions): 73 pass the title rule. Plain research scientist postings include demography and survey science, server demand forecasting, and infrastructure reliability, and AI in a title also names marketing and business development roles; the rule drops those. The 73 left are AI roles, a few of them borderline (security for AI, analytics for AI ranking, AI hardware bring-up). 90 of 91 state a range in USD."  }}

In and out

Part 6, Figures 15 to 17

Where Meta's AI staff studied and worked before, who moved between Meta and ten employers since June 2025, and the base pay Meta's AI postings state.

15

More of the lab once worked at Google (14.2% against 8.3%); China-educated shares match

Earlier employers and schools as a share of each group; China-educated as a share of those listing a bachelor's. Scale 0 to 30%

Google, including Google DeepMind: 14.2% / 8.3%; Amazon: 11.8% / 10.2%; Microsoft: 9.8% / 7.7%; Apple: 3.8% / 3.0%; ByteDance or TikTok: 1.7% / 2.3%; OpenAI: 1.3% / 0.3%; NVIDIA: 1.1% / 0.7%; University of California, Berkeley: 5.7% / 3.9%; Carnegie Mellon University: 5.4% / 3.8%; Stanford University: 5.1% / 3.9%; Indian Institutes of Technology: 4.5% / 3.9%; University of Illinois Urbana-Champaign: 3.3% / 2.4%; Massachusetts Institute of Technology: 2.4% / 1.3%; Mainland-China bachelor's, of those listing one: 19.8% / 19.6%; Of them: individual contributors: · / 21.8%; Of them: manager and above: · / 15.6%

What it shows

The lab also studied more often at Carnegie Mellon and Berkeley (5.4% and 5.7%, against 3.8% and 3.9%); the Stanford gap is within chance. Among the rest, 21.8% of individual contributors and 15.6% of managers and above are China-educated.

Method and limits

An earlier employer is any past non-intern job there, and a school any degree there; rows overlap. China-educated uses the institution list of the China-educated study. The lab's split by level does not reach 20 in every cell, so it is not shown ("·"). Anthropic is left out: no one in the lab and fewer than 10 in the rest.

Source: Metix AI Platform, profiles, 2026-09-22.

Querymapping.json (population, sources, china_educated)
POST /v1/people/query · queries/mapping.json (population, sources, china_educated)
{  "population": {    "id": "p2",    "label": "Meta's visible AI staff",    "note": "A current Meta entry (the stage-1 entry: experience.company.name match Meta and experience.is_current eq true), an AI term in the headline or current title, and three filters that apply to everyone except the lab. The lab's 936 people (queries/population.json) are all in P2 by construction, so the lab node is exactly the stage-1 population and no stage-2 count of the lab differs from a stage-1 count by a small group. Everyone else also needs a current Meta entry whose employer size is 10,001+, a current_function outside non_technical_functions, and a current title with none of non_technical_title_words.",    "ai_terms": [      "machine learning",      "ML",      "deep learning",      "computer vision",      "NLP",      "natural language processing",      "LLM",      "LLMs",      "large language models",      "reinforcement learning",      "generative AI",      "GenAI",      "artificial intelligence",      "superintelligence",      "MSL",      "FAIR"    ],    "ai_terms_fields": [      "headline",      "current_title"    ],    "title_only_terms": [      "AI"    ],    "employer_size": "10,001+",    "non_technical_functions": [      "Human Resources",      "Sales",      "Marketing",      "Administrative",      "Finance & Accounting",      "Legal",      "Customer Service",      "Real Estate"    ],    "non_technical_title_words": [      "marketing",      "sales",      "recruiter",      "recruiting",      "sourcer",      "sourcing",      "talent",      "administrative",      "assistant",      "counsel",      "attorney",      "paralegal",      "accountant",      "communications",      "partnerships"    ],    "lab_note": "26 of the 936 list a non-technical function (recruiting and administration, mostly) and 12 more a non-technical title word. They stay in P2 as members of the lab, so the lab row carries them (about 4% of it) while the rest of P2 does not; the function table shows them among other functions.",    "decisions": [      "The draft term list (machine learning, AI, artificial intelligence, research scientist, research engineer, applied scientist, deep learning, LLM, computer vision, NLP, superintelligence, MSL, FAIR, generative AI, GenAI) gave 15,229 current Meta profiles and 14,297 after the function filter.",      "AI alone counts only in the current title. 4,003 draft profiles entered only through AI in the headline, where it is mostly a buzzword (AI-driven testing, cloud and AI platforms); in the first slice read, about half of them held AI roles.",      "Research scientist, research engineer, and applied scientist are not AI terms. Matching is word by word, so research engineer also matched titles such as a software engineer in Reality Labs Research, and plain research titles at Meta also cover people research, survey science, software testing, photonics, materials, and physical modeling. With these titles admitted, reads in California and Washington found 75 to 82% in AI roles; without them, 92%. 2,180 profiles entered only through such a title and are left out; research scientists who name an AI term, FAIR, or the lab stay in.",      "The employer must be the 10,001+ company. Outside the US the name Meta also matches small companies: 6 of 25 profiles in one slice outside the US worked at other firms whose names contain Meta, none of them sized 10,001+. The rule drops 845 draft profiles, most of them Meta work recorded under variant names or through staffing agencies, and a few that are not Meta at all.",      "ML, LLMs, large language models, natural language processing, and reinforcement learning were added; together they bring 547 profiles the draft missed.",      "Title words exclude marketing, sales, recruiting, and other non-technical roles whose current_function is missing or not in the excluded set (administrative, marketing, and sales roles in the reads)."    ],    "audit": "365 profiles read in whole slices, each a narrow filter read in full (one state or country and a range of total_experience_months), never the top of a search. The first slice (49 profiles of the draft) and three candidate slices (53) set the rules above. 150 profiles in eight whole slices fall in the final P2: 138 (92.0%) hold AI roles by their title, headline, and current job description. The last slice, 39 profiles read after the definition was fixed, gave 35 (89.7%) when two borderline creative and advisory roles count as misses. The misses are stale profiles that still list a current Meta job, generic engineers whose headline lists AI words, and a few operations roles. Contract data workers (prompt engineers, annotators, knowledge experts) who list Meta count as AI roles; they appear under data titles. No lab member with a non-technical function falls in the slices read. Records read for the audit are kept private.",    "size_note": "The name Meta on a current entry gives 66,757 profiles in the US and more than 100,000 elsewhere, because outside the US it also matches other companies. With the employer size 10,001+ in the same entry the count is exact on both sides. A company's reported headcount is the comparison; visible profiles also include contractors and profiles not yet updated after someone left."  },  "sources": {    "employers_note": "Anyone with a past (not current) non-intern job at the employer, at any time, for the lab and the rest of P2; rows overlap. The first six are the stage-1 ever-worked-at rows, with the same definitions, so the lab's counts equal data/sources.json.",    "employers": [      {        "id": "google",        "label": "Google, including Google DeepMind",        "names": [          "Google",          "DeepMind"        ]      },      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "bytedance",        "label": "ByteDance or TikTok",        "names": [          "ByteDance",          "TikTok"        ]      }    ],    "schools_note": "Any education entry at the school, any degree; rows overlap.",    "schools": [      {        "id": "stanford",        "label": "Stanford University",        "names": [          "Stanford"        ]      },      {        "id": "carnegie-mellon",        "label": "Carnegie Mellon University",        "names": [          "Carnegie Mellon"        ]      },      {        "id": "berkeley",        "label": "University of California, Berkeley",        "names": [          "Berkeley"        ]      },      {        "id": "mit",        "label": "Massachusetts Institute of Technology",        "names": [          "Massachusetts Institute of Technology"        ]      },      {        "id": "uiuc",        "label": "University of Illinois Urbana-Champaign",        "names": [          "Urbana"        ]      },      {        "id": "iit",        "label": "Indian Institutes of Technology",        "names": [          "Indian Institute of Technology"        ]      }    ]  },  "china_educated": {    "institutions_file": "cases/china-educated-ai-talent-2026/queries/institutions.json",    "note": "China-educated means a Bachelor entry that names an institution on the China-educated study's mainland list (its match and exact names, none of its exclude words), read from that file. The query sends the match and exact names in one in leaf and the exclude words in another; on P2 it gives the same 1,227 profiles as the study's own form with one match leaf per name. The share is of people with any Bachelor entry.",    "level_gate_count": 20,    "level_note": "By level (individual contributors, that is Intern, Specialist, and Senior, against manager and above) only where all four cells of a group (China-educated and Bachelor-listed, at each level) hold 20 or more."  }}

16

Since June 2025, 191 left Meta for OpenAI and 37 joined from it; 691 joined from Amazon

People who moved between Meta and each employer since June 2025, every role. Both sides share one scale

OpenAI: 191 left, 37 joined; Anthropic: 81 left, <10 joined; Google DeepMind: 58 left, 41 joined; xAI: 33 left, 11 joined; Thinking Machines Lab: 16 left, <10 joined; Amazon: 153 left, 691 joined; Microsoft: 154 left, 343 joined; Google, other than DeepMind: 308 left, 317 joined; Apple: 134 left, 158 joined; NVIDIA: 82 left, 20 joined

What it shows

379 people left Meta for the five AI labs and at most 107 joined from them. With Amazon, Microsoft, and Apple it runs the other way: 691 joined from Amazon and 153 left for it, and 343 joined from Microsoft against 154. NVIDIA is the exception: 82 left for NVIDIA and 20 joined from it. Google outside DeepMind is about even (317 in and 308 out). Across all employers, 8,054 left and 7,787 joined.

Method and limits

A move counts when the new job started on or after June 1, 2025 and the previous one ended on or after May 1, 2025, internships excluded, for every role at Meta, not only AI roles. Most who moved to the AI labs carry no AI words in their title, which fits labs that title staff members of technical staff. Profile dates stop in April 2026, so later moves, the May 2026 reduction among them, are not counted; people cut from Meta's AI teams in October 2025 count once their next job appears. A "<10" is drawn as an outline at the largest value it could hide.

Source: Metix AI Platform, profiles, 2026-09-22.

Querymapping.json (flows)
POST /v1/people/query · queries/mapping.json (flows)
{  "flows": {    "note": "Moves since 2025-06-01, with symmetric windows, non-intern entries only. Meta is the same employer as in P2: the name Meta with employer size 10,001+, since outside the US the name alone also matches other companies; Facebook, Instagram, WhatsApp, and Oculus entries count by name. Out to X: a current entry at X that started on or after 2025-06-01, a Meta entry that ended on or after 2025-05-01, and no current entry under the name Meta at any size, so a current Meta job whose entry lacks a size never counts as a move away. In from X: a current Meta entry that started on or after 2025-06-01, an entry at X that ended on or after 2025-05-01, and no current entry at X. All movers: the same with any employer outside the Meta names. The AI-titled subset adds the P2 AI terms in the headline or current title, and is counted only where the row holds 10 or more. Rows by employer can overlap.",    "start_from": "2025-06-01",    "ended_from": "2025-05-01",    "peers": [      {        "id": "openai",        "label": "OpenAI",        "names": [          "OpenAI"        ]      },      {        "id": "google-deepmind",        "label": "Google DeepMind",        "names": [          "DeepMind"        ]      },      {        "id": "anthropic",        "label": "Anthropic",        "names": [          "Anthropic"        ]      },      {        "id": "xai",        "label": "xAI",        "names": [          "xAI"        ]      },      {        "id": "thinking-machines",        "label": "Thinking Machines Lab",        "names": [          "Thinking Machines"        ]      },      {        "id": "microsoft",        "label": "Microsoft",        "names": [          "Microsoft"        ]      },      {        "id": "apple",        "label": "Apple",        "names": [          "Apple"        ]      },      {        "id": "amazon",        "label": "Amazon",        "names": [          "Amazon",          "AWS"        ]      },      {        "id": "nvidia",        "label": "NVIDIA",        "names": [          "NVIDIA"        ]      },      {        "id": "google",        "label": "Google, other than Google DeepMind",        "names": [          "Google"        ],        "exclude": [          "DeepMind"        ]      }    ],    "limits": "Profiles lag: start dates stop in April 2026, so a move needs a new entry by then to count, and the most recent months are undercounted. Meta cut more than 600 roles in its AI organization in October 2025 (SiliconANGLE, 2025-10-22, https://siliconangle.com/2025/10/22/meta-lays-off-600-ai-workers-looks-streamline-superintelligence-labs/); people who left then count as out-movers only once their next job appears. An entry whose date gives only a year reads as January of that year."  }}

17

The median top of Meta's posted AI base ranges falls between $250,000 and $295,000

The top of the stated base range in US postings, by band; each row is 100%

All AI postings (197): Under $215k 18, $215k to $250k 59, $250k to $295k 75, $295k to $340k 37, $340k and up 8; Research scientist (33): Under $215k 3, $215k to $250k 12, $250k to $295k 13, $295k to $340k 3, $340k and up 2; Machine learning or software engineer (99): Under $215k 8, $215k to $250k 24, $250k to $295k 41, $295k to $340k 22, $340k and up 4

What it shows

Of the 197 US AI postings that state pay, the median start of the range falls between $180,000 and $215,000, and the median top between $250,000 and $295,000. Research scientist postings do not top out higher: 54.5% reach $250,000 or more, against 67.7% of machine learning and software engineer postings.

Method and limits

US postings in dollars that state both ends; base pay as posted, without bonus, equity, or benefits, and nothing here describes what anyone is paid. A family is shown only with 15 or more such postings; research engineer (12) is not.

Source: Metix AI Platform, jobs, 2026-09-22.

Querymapping.json (postings, pay)
POST /v1/jobs/query · queries/mapping.json (postings, pay)
{  "postings": {    "note": "Meta postings (company.name match Meta; the jobs index has no employer size, so a few postings at other companies named Meta may be included) posted in the last 180 days (posted_date gte now-180d) whose title contains a P2 AI term or AI, and none of the non-technical title words. The index holds open postings only, and every one read was posted within three weeks of the snapshot, so this is the demand open on the day. A role posted in several cities counts once per city. Counts of postings carry no minimum.",    "posted_from": "now-180d",    "families": [      {        "id": "research-scientist",        "label": "Research scientist",        "title": [          "research scientist"        ]      },      {        "id": "research-engineer",        "label": "Research engineer",        "title": [          "research engineer"        ]      },      {        "id": "ml-software-engineer",        "label": "Machine learning or software engineer",        "title": [          "software engineer",          "machine learning engineer",          "ML engineer",          "AI engineer"        ]      },      {        "id": "data",        "label": "Data",        "title": [          "data",          "annotation",          "annotator",          "labeling",          "prompt",          "evaluator"        ]      },      {        "id": "product-program",        "label": "Product, program, and design",        "title": [          "product manager",          "product management",          "program manager",          "designer"        ]      }    ],    "families_note": "Exclusive, first match in this order; the rest are other (hardware, production, security, network, and leadership titles).",    "audit": "91 read (see directions): 73 pass the title rule. Plain research scientist postings include demography and survey science, server demand forecasting, and infrastructure reliability, and AI in a title also names marketing and business development roles; the rule drops those. The 73 left are AI roles, a few of them borderline (security for AI, analytics for AI ranking, AI hardware bring-up). 90 of 91 state a range in USD."  },  "pay": {    "note": "US postings (location.country eq United States) in USD that state both salary.annual_min and salary.annual_max, by family. A family is published only with 15 or more such postings. Counts at thresholds describe the posted base range; bonus, equity, and benefits are not in these figures, and no statistic describes what anyone is paid.",    "gate_count": 15,    "annual_min_thresholds": [      150000,      180000,      215000,      265000    ],    "annual_max_thresholds": [      215000,      250000,      295000,      340000    ],    "thresholds_note": "Meta's posted ranges fall on a few fixed bands (for example 122,000 to 181,000, 154,000 to 217,000, 184,000 to 257,000, 219,000 to 301,000, and 271,000 to 347,000 in the reads), so these thresholds separate the bands."  }}

Run it

Three ways in. Each says what it costs before you start.

1 API Credit buys 25 search results or 5 full records; $1 buys 30 API Credits.

Run it in your agent

500 to 600 API Credits$16.67 to $20.00More than the 100 free API Credits

Your agent stops and asks before spending more than 650 API Credits.

Your agent follows the prompt step by step: it reads the rules, runs the counts, checks the definitions the prompt asks it to check, and writes the files and charts. Use an agent that can write files, such as Claude Code or Codex.

Set up onceKey, connection, and a free check. Skip this if your agent already reaches the Platform.
  1. 1Get a key

    Create a key on the Metix AI Platform →

    New accounts get 100 API Credits once, valid for 30 days. Set the key in the shell you start your agent from, or add the line to ~/.zshrc or ~/.bashrc so every new terminal has it:

    Shell
    export METIX_KEY=metix_xxxxxxxx
  2. 2Connect your agent

    Claude Code

    Registers the Platform for every project. Start claude in any folder and the ten metix tools are there.

    Shell
    : "${METIX_KEY:?run step 1 first}" &&
    claude mcp add --scope user --transport http metix \
      https://mira-api.metix.ai/mcp \
      --header "Authorization: Bearer $METIX_KEY"
    MCP setup guide →

    Codex

    Registers the same server. The key stays in your environment instead of the config file.

    Shell
    codex mcp add metix \
      --url https://mira-api.metix.ai/mcp \
      --bearer-token-env-var METIX_KEY
    MCP setup guide →

    Skills

    Four skills that teach any agent the Platform's endpoints and query rules, for agents without MCP. The installer starts with none ticked: press space on each, then enter.

    Shell
    npx skills add MetixAI-Official/metix-skills
    Skills install guide →

    Other MCP

    Point the client at this endpoint over streamable HTTP with both headers; without the Accept header the server answers 406. Older clients use /sse on the same host.

    Endpoint and headers
    https://mira-api.metix.ai/mcp
    Authorization: Bearer <your key>
    Accept: application/json, text/event-stream
    MCP setup guide →
  3. 3Check the setup

    Ask this first. It reads your balance and the field list, runs no search, and costs nothing:

    Prompt for your agent
    Use the Metix AI Platform to check my key status and read the contract; both are free. Then tell me my API Credit balance and which datasets I can query. Do not run any search.

4Paste the prompt

Start your agent in an empty folder, then paste. It writes its files there.

The question

Answer two questions with the Metix AI Platform: of the people who name Meta Superintelligence Labs on their public profile, how many were already at Meta when the lab formed in June 2025, and where did the rest come from? And where do Meta's visible AI staff sit, and where does the lab sit among them? Work only through the public Platform (REST at https://mira-api.metix.ai, the MCP server, or the metix-skills) with the key in METIX_KEY, and never print the key.

  1. 01Read before querying

    Call GET /contract (free) and build every condition from querySpecByEntity.profile and querySpecByEntity.job; read GET /docs/api/people (free) too. Conditions about one job go inside one has_experience entry, so they describe the same job. Text fields match word by word, in any order, ignoring case; current_function, current_seniority, experience.seniority, and experience.company.type take fixed values with eq or in. A query holds at most 64 conditions and nests at most 6 levels. A count with size 1 costs 1 API Credit and nothing when it finds no one, a search 1 API Credit per 25 IDs, and a detail read 1 API Credit per 5 records, so plan every published number as a count. Send each query once with an extra word that matches nothing first: the Platform checks it against its limits and charges nothing for an empty result. Check the balance with GET /auth/key/status (free) at the start and at the end, and stop and ask before the run passes 650 API Credits.

  2. 02Population

    A current job at Meta (experience.company.name match "Meta" and experience.is_current eq true, in one has_experience entry) and a headline or current_title that matches superintelligence or MSL. Check with counts that company.name eq "Meta" gives the same total, so the name catches no other company, and that Facebook is the same employer on the Platform rather than a separate test. An earlier Meta job is an entry at Meta, Facebook, Instagram, WhatsApp, or Oculus. An internship is an entry with experience.seniority eq "Intern" or an experience.title that matches intern, tested in the same entry. Count the narrower definition (superintelligence only) too, and, as context, current Meta profiles that name FAIR but no lab term.

  3. 03Audit before counting

    Search ranks results by match quality, so the top of a search looks cleaner than the group. Read whole slices instead: a search with size 25 costs 1 API Credit and returns the total, so a slice of 25 or fewer comes back complete. Read about 60 profiles in five slices (people with no Meta entry before June 2025 in one state, people with one in two cities, the group whose Meta entry has no start date, and people who write MSL but not superintelligence in one state) with POST /entity/v1/profiles/detail-by-id and _source headline, current_title, current_function, and experience company.name, title, start_date, end_date, is_current, and seniority, never a name. Report the share that describe working in the lab, check that the veterans had real Meta jobs rather than only internships, and look for Meta company names the list misses. Read the matches for each sub-team term, a whole slice of the Microsoft group, and every open posting the same way. Keep every record out of every published file.

  4. 04Cohorts

    The cut is 2025-06-01. Count the total; people with an earlier Meta entry that started before the cut; the same without internships; and people who stayed, with a non-intern Meta entry that started before the cut and is current or has experience.end_date gte "2025-05-01" (the any of those two inside the entry). Returned is the non-intern count minus stayed, former interns the first count minus the non-intern count, and new to Meta the total minus the first count; the last three together are the outside hires. Count total, earlier entry, and stayed for the narrower definition too, and the four groups for current Microsoft profiles that name superintelligence, with a cut of 2025-11-01 and an end of 2025-10-01.

  5. 05Dates

    Count people with an earlier Meta entry that started before each of 2012-01-01, 2014-01-01, 2016-01-01, 2018-01-01, 2020-01-01, 2022-01-01, 2023-01-01, 2024-01-01, 2025-01-01, the cut, 2025-07-01, 2025-08-01, 2025-09-01, 2025-10-01, 2026-01-01, and 2026-04-01; the difference between two dates is a band of first-join dates. Repeat without internships up to the cut. For people who stayed, count those with a current Meta entry that started on or after each of 2025-01-01, 2025-04-01, 2025-06-01, 2025-07-01, 2025-08-01, 2025-09-01, 2025-10-01, 2026-01-01, and 2026-04-01, which places each by their newest role. For Microsoft, count before 2012, 2016, 2020, 2023, 2025-01-01, and its cut. Profiles lag, so count Meta entries by start month across all profiles and say which month is the last complete one.

  6. 06Sources

    For the outside hires (the population, not stayed), count a job entry with experience.end_date gte "2025-01-01" that is not an internship, at each employer in this order: OpenAI, DeepMind, Anthropic, xAI, Thinking Machines, Apple, Scale AI, Google, Microsoft, Amazon or AWS, NVIDIA; then experience.company.type eq "Educational", where internships count. Each row leaves out anyone an earlier row matched, and the rest are other employers or none. For the whole group, count anyone with a past non-intern job at Google or DeepMind, OpenAI, Anthropic, Apple, Microsoft, and Amazon or AWS; those rows overlap.

  7. 07Roles, teams, places, and postings

    Count current_function eq Research, Engineering and Technical, and Product, and current_function exists, for the group and for the outside hires. For people whose current function is Research, in the lab and in the baseline of current Meta staff with that function and no lab term, count current_seniority in Senior, in Manager or any level above it, and exists, and a Doctorate education entry; count doctorates for the group and the outside hires as well. Count headline or title mentions of FAIR, TBD, infra or infrastructure, and PAR or product applied research. Count location.country United States, location.state California, New York, and Washington, Bay Area cities within California, and the United Kingdom. For open Meta postings (is_open eq true, company.name match "Meta", and a title matching superintelligence or MSL or a description matching superintelligence labs), count the total, research titles, Menlo Park, San Francisco, New York, postings that state min_experience_months and those at 36 or less, the seniority label, postings since 2026-09-01, and every open Meta posting.

  8. 08Outputs

    Write aggregate files with "unit" ("profiles" or "jobs"), the snapshot date, and the query files they came from. Every count of people from 1 to 9 is written "<10". A band of dates that would hold 1 to 9 people merges with a neighbour, never across the cut. When the hidden cells of a published sum add up to 1 to 9, withhold one more cell. Split the narrower definition and the Microsoft group only as finely as keeps every difference between published counts at 0 or at least 10. Publish a share only when its denominator is at least 30. Check every difference a reader could form before writing anything, and name no individual in any text you write, including people named in the press; linking to press coverage is fine.

  9. 09Charts

    The first-join bands as a timeline with the cut marked, the months after it for people new to Meta, and Microsoft on the same axis; the four groups as one bar, with the narrower definition and Microsoft beneath it; outside hires by where they had just worked, with the five AI labs grouped; current function for people who stayed and outside hires; doctorates and levels of the lab's research staff against the baseline; sub-team mentions; states; open postings by place. Label bars directly, and give each chart a title that states its finding in neutral words.

  10. 10Limits

    The population is people who chose to name the lab, not the lab: people who write only FAIR or nothing are missing, and newcomers and long-time staff may name it at different rates. Counts are visible profiles, never a headcount, and since stale profiles count as current, not a guaranteed floor either. Some profiles are out of date, and recent months are undercounted. A date given only as a year reads as January, which moves some people across the cut; say in which direction. A new Meta entry is a lower bound on who moved, since many team changes add none. Sources rest on end dates. The Microsoft group is small and includes recruiters. Postings are one day. Use neutral verbs (join, leave, move, hire) and describe no pay other than the posted base ranges of step 17.

  11. 11The map's population

    Meta's visible AI staff (P2) are people with the stage-1 current Meta entry and an AI term in the headline or current title: machine learning, ML, deep learning, computer vision, NLP, natural language processing, LLM, LLMs, large language models, reinforcement learning, generative AI, GenAI, artificial intelligence, superintelligence, MSL, or FAIR, or AI in the current title alone. Everyone outside the lab also needs a current Meta entry with experience.company.size eq "10,001+" in the same has_experience, a current_function outside Human Resources, Sales, Marketing, Administrative, Finance & Accounting, Legal, Customer Service, and Real Estate, and a current title without marketing, sales, recruiter, recruiting, sourcer, sourcing, talent, administrative, assistant, counsel, attorney, paralegal, accountant, communications, or partnerships. Put each of those three filters in an any beside the two lab-term leaves, so every stage-1 lab member stays in P2 and the lab is the same group in both stages. Send a list of terms with in on a text field: it matches like one match leaf per term and keeps each query under 64 conditions.

  12. 12Audit the map

    Read at least 100 profiles in whole slices of P2 (one state or country and a narrow range of total_experience_months, every match read) with headline, current_title, current_function, summary, and the current entry's title, description, company name, and size, never a name, and report the share in AI roles; it must reach 90%. Test the traps the reads found before settling the terms: AI alone in a headline is mostly a buzzword; research scientist and research engineer match word by word and cover people research, survey science, photonics, and software engineers in Reality Labs Research; outside the US the name Meta also matches other companies. Read whole slices of the unit terms as well (GenAI, generative AI, AR and VR words, infra) and decide from them whether each names a team or a topic, and in which order the units go.

  13. 13Units and size

    Count current Meta profiles in the US, and current entries at the 10,001+ Meta in the US and elsewhere, so that no count comes back as 100000+. Put each P2 member in the first unit whose terms appear in the headline or current title, in this order: the lab (superintelligence, MSL); FAIR; Reality Labs, with AR, VR, XR, augmented reality, virtual reality, mixed reality, wearables, smart glasses, and Oculus; infra or infrastructure; ads, monetization, or advertising; ranking, recommendation, recommendations, recommender, or recsys; integrity; Instagram, WhatsApp, Messenger, or Threads; GenAI; and none of these. For each unit and for P2, count the total, the Research and the Engineering and Technical functions, manager and above, a doctorate, the US, a Bachelor entry, and a Bachelor entry at a mainland-China institution on the list in cases/china-educated-ai-talent-2026/queries/institutions.json. The lab's total, functions, doctorates, and US count are the stage-1 counts of the same group.

  14. 14Functions, levels, and titles

    For the lab and for P2, count data titles (data, annotation, annotator, labeling, prompt, knowledge expert, rater, evaluator) outside Research, then Engineering and Technical, Product with Design and Project Management, and no function, each without a data title; research is the Research count of step 13 and other is what is left. Count current_seniority Intern, Specialist, Senior, Manager, and Head or Director; Vice President and above is manager and above minus Manager, Head, and Director. Count current titles containing machine learning engineer, software engineer, research scientist, research engineer, engineering manager, product manager, program manager, data scientist, data engineer, prompt engineer, production engineer, and director, and keep the ten largest.

  15. 15Directions

    For foundation models and LLMs, post-training and alignment, multimodal and perception, AI infrastructure, agents, ranking and recommendations, AR, VR, and devices, and safety and evaluation, count the lab and P2 by a phrase in the headline or current title or a single word in skills, and Meta AI postings (step 17) by a phrase in the title or a single word in the description. Every Meta description carries a paragraph on augmented and virtual reality and most mention responsible AI, and a phrase matches its words anywhere in a long text, so read a whole slice of postings before choosing the description words; the list is in queries/mapping.json.

  16. 16Sources and moves

    For the lab and P2, count a past non-intern job at Google or DeepMind, OpenAI, Anthropic, Apple, Microsoft, Amazon or AWS, NVIDIA, and ByteDance or TikTok, and any education entry at Stanford, Carnegie Mellon, Berkeley, the Massachusetts Institute of Technology, Urbana, and an Indian Institute of Technology. Split the China-educated share by level (Intern, Specialist, and Senior against manager and above) only where every cell holds 20 or more. For moves since 2025-06-01, with the same window on both sides and Meta the same employer as in P2 (the name with employer size 10,001+; Facebook, Instagram, WhatsApp, and Oculus by name): out to X is a current non-intern entry at X that started on or after 2025-06-01, a non-intern Meta entry that ended on or after 2025-05-01, and no current entry under the name Meta at any size; in from X is a current non-intern Meta entry that started on or after 2025-06-01, a non-intern entry at X that ended on or after 2025-05-01, and no current entry at X. Count OpenAI, DeepMind, Anthropic, xAI, Thinking Machines, Microsoft, Apple, Amazon or AWS, NVIDIA, Google without DeepMind in the same entry, and any employer outside the Meta names, and each again with the AI terms where the row holds 10 or more.

  17. 17Postings and pay

    Take Meta postings with posted_date gte "now-180d" whose title holds a P2 AI term or AI and none of the non-technical title words, and count them by family, first match in order: research scientist; research engineer; software engineer, machine learning engineer, ML engineer, or AI engineer; data titles; product manager, product management, program manager, or designer; other. For US postings in USD that state both salary.annual_min and salary.annual_max, count all of them and each of the first three families, and for each with 15 or more, count annual_min at or above 150,000, 180,000, 215,000, and 265,000 and annual_max at or above 215,000, 250,000, 295,000, and 340,000. That is the posted base range, not pay.

  18. 18Map outputs

    Write one aggregate file per measure: data/map_size, map_units, map_functions, map_levels, map_titles, map_directions, map_sources, map_flows, map_postings, and map_pay. Every count of people from 1 to 9 is "<10" and a share needs a base of 30. List every sum a reader can form (the units add up to P2, each table row to its total, a unit's US staff sit inside its count, a stage-2 lab cell sits inside the stage-1 cell that holds it, a move row's AI-titled part sits inside the row), and withhold the smallest cell until no group of 1 to 9 people can be worked out. Stage 1 published the lab's functions, so when splitting out data titles would leave such a difference, merge the lab's other and none. Write nothing if any check fails.

  19. 19Map charts and limits

    The org map as a tree: P2, its units, and the lab with its stage-1 groups beneath it; then units by function and level, directions as profiles against postings, sources, moves as paired bars, and posted pay bands by family. Say that units are what people write, not Meta's org chart; that P2 counts profiles, not staff, with the accuracy the audit measured; that postings are one day's open roles and a role posted in several cities counts once per city; that moves rest on profiles that lag to April 2026 and fall across the October 2025 reductions in Meta's AI organization; and that hiring difficulty is out of scope for one company.

What you get

The aggregate files and the chart, a note on what the audits found and what they changed, and the API Credits the run spent, read from the balance before and after.

A call that returns 402 insufficient_quota means the key works and the balance is empty.

Reproduce the numbers

353 API Credits$11.77More than the 100 free API Credits

A short standard-library Python script sends the committed queries as counts and writes the aggregate files this page is built from. It needs Python and METIX_KEY set in the shell (step 1 of the agent path); your agent can run these lines for you as well.

Shell
curl -fsSL https://platform.metix.ai/casebook/source/meta-superintelligence-labs-2026.tar.gz | tar xz
cd meta-superintelligence-labs-2026
: "${METIX_KEY:?set METIX_KEY first}" && python3 cases/meta-superintelligence-labs-2026/fetch.py

What you get

data/*.json and data/receipt.json. Run git diff cases/meta-superintelligence-labs-2026/data to see what moved: the numbers should match, apart from what changed in the data since the snapshot.

Adapt it

Costs what your version reads. Write your own ceiling into step 1 of the prompt.

The prompt is the case. Change the parts in this table and your agent answers your question instead, with the same checks and the same way of reporting cost.

To change Edit For example
The lab The employer and lab terms in step 2, audited again in step 3 Google DeepMind staff who name Gemini, or Microsoft AI
The date the lab formed The cut in steps 4 and 5 2025-11-01 for Microsoft's superintelligence team
Where newcomers came from The employer order in step 6 Add Mistral AI or Cohere
The map's units The order and terms in step 13, audited again in step 12 A unit for WhatsApp alone, or Llama
The API Credit ceiling Steps 1, 3, and 12 Run steps 1 to 10 alone for about 200 API Credits

What to ask before running it

When someone brings a looser version of this question, settle these first. Each one changes the query or the cost:

  1. Which words mark someone as part of the lab, and does any of them also mean something else?
  2. Which company names count as the same employer, now and in the past?
  3. What date separates people who were there from people who joined?
  4. Do internships count as having been there?
  5. How many API Credits may the audit reads spend?
  6. Which words in a headline name a team, and which only a topic?

Method and limits

How the population was defined, counted, and checked, and what the numbers cannot show.

Population

Profiles on the Metix AI Platform on September 22, 2026 with a current job at Meta whose headline or current title says superintelligence or MSL, matched word by word and ignoring case: 936 people, 790 of whom write superintelligence. Definitions: queries/population.json and queries/measures.json.

Four groups

The cut is June 1, 2025; Meta announced the lab at the end of that month. At Meta when it formed: a Meta job, not an internship, that began before the cut and was current or ended in May 2025 or later. Returned: a Meta job, not an internship, before the cut, none of which lasted into May 2025. Former interns: only internships at Meta before the cut. New to Meta: no Meta entry before the cut, including fewer than 10 whose Meta entry has no start date. The last three are the outside hires. An earlier Meta job is an entry at Meta, Facebook, Instagram, WhatsApp, or Oculus; an internship is an entry whose seniority is Intern or whose title says intern.

Where outside hires came from

A job entry that ended in 2025 or later, at the first match in a fixed order. Internships do not count for companies, so a doctoral student's summer at a lab is not where they came from; they do count for universities, because doctoral students' entries often carry the Intern label.

Dates

A first join is the start of a person's earliest Meta entry, taken from the counts of people with a Meta entry before each date, so no one is read. A period that would hold 1 to 9 people merges with its neighbour. Profiles lag: across all Meta entries on the Platform, 3,988 start in 2026, 1,104 of them in March and April and fewer than 10 in May or later, so dates stop in April 2026 and the last months are undercounted. A date given only as a year reads as January.

Audits

The definition carries the study, so 63 profiles were read in whole slices, never the top of a search: 62 describe working in the lab and the rest recruit for it. The reads changed the definition three ways: MSL counts too (13 of 14 read use it for the lab), Instagram, WhatsApp, and Oculus join Meta as earlier employers, and the few data labeling and red teaming roles are counted apart. No record read is published.

Limits

This is people who chose to name the lab, not the lab: people in its groups who write only FAIR or nothing are missing, and if newcomers name it more readily than long-time staff, or less, the share moves with them. Counts are of visible profiles, not staff: people who do not name the lab are missing and profiles left out of date remain, so a count is neither a headcount nor a guaranteed floor (figure 11 finds 1.7 current Meta profiles for each employee Meta reports). A date given only as a year reads as January, so a newcomer who wrote only 2025 counts among those at Meta when the lab formed, and someone who stayed can count among those who returned. Where outside hires came from rests on end dates, and 47.2% of them have no job at the listed employers or a university that ended in 2025 or later. The Microsoft group is small and includes recruiters, and profile lag hides more of its months than of Meta's. Postings are one day. No one is named, and nothing here describes pay.

Stage 2: who is counted

Profiles with a current job at Meta, the 10,001+ company, whose headline or current title names an AI term (listed in queries/mapping.json; AI alone counts only in the title), outside non-technical functions and titles: 8,257 people. The lab's 936 are included whole, so the lab matches stage 1; about 4% of them list a non-technical function or title. Research titles that name no AI term are left out, because at Meta they also cover people research, survey science, and hardware. Reads of 150 profiles in eight whole slices found 92% in AI roles.

Stage 2: units, work, and directions

A unit is the first of the lab, FAIR, Reality Labs, infrastructure, ads, ranking, integrity, apps, and GenAI whose words appear in the headline or title. Function and level are the Platform's fixed values for the current job; a title counts when it contains a candidate's words. Directions are multi-label: staff are matched on headline, title, and skills, and Meta's 218 AI postings of the last 180 days on title and description.

Stage 2: flows and pay

A move counts when the new job started on or after June 1, 2025 and the previous one ended on or after May 1, 2025, internships excluded, for every role at Meta. Pay is the base range stated in US postings in dollars, shown for a title family with 15 or more such postings.

Limits of stage 2

Current Meta profiles outnumber Meta's reported headcount 1.7 to 1, so every stage-2 count is of profiles, and none is a floor or a ceiling on staff. Units are what people write, not Meta's org chart, and 62.8% name none of the nine units searched. Profile dates stop in April 2026, so the May 2026 reduction and later moves are not in the data. Levels are a visible hierarchy from titles, not reporting lines. How hard roles are to fill is out of scope: it needs a comparison across employers, not one company.

Sources

Meta's announcement of the lab (CNBC, June 30, 2025); Meta's investment in Scale AI (Axios, June 12, 2025); Microsoft's superintelligence team (CNBC, November 6, 2025); the lab's four teams (Built In, August 27, 2025); cuts in Meta's AI teams (SiliconANGLE, October 22, 2025); Meta's headcount (SEC filing, July 29, 2026).

The last reproduction

What reproducing this case cost the last time the script ran, read from the Platform's own balance before and after.

Ran on
2026-09-22
Calls
358
Search results
353
Records read
0
API Credits
353

Search results are IDs returned by searches, one per count query and one per match on a full search. Records are postings or profiles read in full: this case reads none.

Making this case cost about 1,629 API Credits more: the agent's audits, trial queries, and earlier runs that the published replay replaced. You do not pay that again.