Google, Meta and Microsoft have each moved a named AI research group under a unit built to make models: Google’s Brain team in April 2023, Meta’s FAIR in June 2025 and Microsoft’s AI Frontiers in September 2026, according to the companies and the people involved.
By our count, that is three reorganizations in about 41 months, from Google’s announcement on April 20, 2023 to the move John Langford dated to “early September” 2026. The latest is the one Langford wrote about on November 6, in a post we covered on Friday.
Business Pill · Vertical Integration
An evergreen Business Engineer explainer on vertical integration. It teaches the general idea only and does not discuss Google, Meta, Microsoft or their research groups.
The key insight:Google merged its Brain team into Google DeepMind in April 2023, Meta put FAIR inside Meta Superintelligence Labs in June 2025, and Microsoft moved the roughly 50-person AI Frontiers group into Microsoft AI in September 2026, according to the companies, CNBC, Reuters and John Langford.
On April 20, 2023, Google said in a post on its blog that “This group, called Google DeepMind, will bring together two leading research groups in the AI field”: the Brain team from Google Research, and DeepMind (Google, The Keyword).
The post said Demis Hassabis, as CEO of Google DeepMind, would lead the development of its most capable general AI systems, “research that will help power the next generation of our products and services”. Jeff Dean took the role of Google’s Chief Scientist, the post said.
Google’s Brain team (April 2023), Meta’s FAIR (June 2025) and Microsoft’s AI Frontiers (September 2026) each moved under a unit that builds models, about 41 months from first to third by our count. Sources: Google, Microsoft, CNBC, Reuters, Meta, hunch.net; dates as published.
Microsoft: A New Unit, Then a Merger
On March 19, 2024, Microsoft chief executive Satya Nadella wrote that Mustafa Suleyman and Karén Simonyan were joining Microsoft to form a new organization called Microsoft AI, “focused on advancing Copilot and our other consumer AI products and research” (Official Microsoft Blog). Suleyman became EVP and CEO of Microsoft AI, the post said.
John Langford wrote on November 6, 2026 that AI Frontiers, a group of about 50 people inside Microsoft Research, was “reorged into Microsoft AI” in early September (hunch.net, The End of AI Frontiers). The post concludes: “So AI Frontiers is over”.
Langford also wrote: “I am just not comfortable being at a company where the research I do has no path to real use”. That is Langford’s account; we found no Microsoft statement responding to it. Microsoft AI’s own plans are in our earlier piece on its models.
Meta: FAIR Moves Under Superintelligence Labs
On June 30, 2025, Mark Zuckerberg announced to staff the creation of Meta Superintelligence Labs, led by Alexandr Wang and Nat Friedman, CNBC reported from an internal memo (CNBC, via NBC New York). CNBC reported that the unit would house Meta’s teams working on foundation models, products and Fundamental Artificial Intelligence Research (FAIR) projects.
On October 22, 2025, Meta said it was cutting around 600 positions out of several thousand roles in Superintelligence Labs, Reuters reported (Reuters). Reuters reported that the cuts would affect FAIR, product-related AI and AI infrastructure teams, and that the TBD Lab building Meta’s next-generation models would not be affected.
On November 19, 2025, Yann LeCun, who joined Meta in 2013 to create FAIR, announced plans to leave at the end of the year to launch a startup, Reuters reported (Reuters). “The creation of FAIR is my proudest non-technical accomplishment”, LeCun wrote. We covered LeCun’s departure and Meta’s current structure.
On April 8, 2026, Meta announced Muse Spark, which it called the first in a new series of large language models built by Superintelligence Labs (Meta Newsroom). “Over the last nine months, Meta Superintelligence Labs rebuilt our AI stack from the ground up”, Meta said.
The Pattern in the Companies’ Own Words
Each announcement ties research to products in its own wording. Google described its merged unit’s work as research to power its products. Microsoft described its new unit as focused on Copilot and other consumer AI products and research. Meta described Superintelligence Labs as the builder of its new model series.
Two of the people who created these groups later said they were leaving: LeCun, who created FAIR, and Langford, whose post says “we organized to create the AI Frontiers sub-organization within MSR” in 2023.
Our Numbers
By our count from the sources above, three of the three companies moved a named research group under a model-building unit between April 2023 and September 2026: Google Brain, FAIR and AI Frontiers. That is about 41 calendar months from the first to the third, our arithmetic. Meta’s cut of around 600 roles came about four months after Superintelligence Labs was formed, also our arithmetic.
The Structural Read
As we read the three announcements, each company described the new unit by what it builds: Google by its general AI systems and products, Microsoft by Copilot and its consumer AI products, and Meta by its Muse series of models.
As we read the dates, the moves came in a sequence rather than at once: one in 2023, one in 2025 and one in 2026, about 41 months from first to third by our count.
As we read Langford’s post, the account from inside one group is about direction, not money: it describes a scaling-focused unit and research with what Langford calls no path to real use.
John Langford, hunch.net, 6 November 2026
“So AI Frontiers is over”
Three Implications
FOR RESEARCHERS Langford wrote that there will not be internships with the folks there in the coming year.
FOR META Reuters reported that Meta’s October 2025 cuts of around 600 roles affected FAIR, while the TBD Lab building its next-generation models was not affected.
FOR MICROSOFT Nadella’s March 2024 post described Microsoft AI as focused on advancing Copilot and its other consumer AI products and research.
The analysis says: “Most product organizations are built to ship for the curve they can see today.”
As we read it, each reorganization here placed a research group inside an organization that ships models and products. The analysis does not discuss Google Brain, FAIR or AI Frontiers.
What Is Not Established
The companies did not describe these moves as closing research. Google called its change a bringing-together of two research groups, and Reuters reported that Superintelligence Labs includes the FAIR team. Only Langford’s post says a group is “over”, and it is one founder’s account. We count only groups named in these sources; other reorganizations may exist. This piece makes no forecast.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
In about 41 months, Google, Meta and Microsoft each moved a named AI research group under a unit charged with building models and products, by our count from the companies’ own statements and reporting. The pattern is consistent across all three: the research group became part of the modelorganization. Our earlier analysis of the scaling schism covers the researchers’ side of the argument.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. Dates, descriptions and quotes are from Google’s April 2023 blog post, Satya Nadella’s March 2024 post, Meta’s April 2026 newsroom post, John Langford’s November 2026 post on hunch.net, and reporting by CNBC and Reuters. The month counts and the phrase model-building unit are our summary and arithmetic. This is not investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
