Meta has employed five founders from Thinking Machines Lab, including an engineer who is said to be worth $1.5 billion.

      In conclusion, Meta has recruited five founding members from Thinking Machines Lab, the AI startup established by former OpenAI CTO Mira Murati, who declined a reported $1 billion acquisition offer. The highest-priced hire was co-founder Andrew Tulloch, who supposedly received a $1.5 billion package over six years. This talent acquisition is part of a larger restructuring initiative in AI, which also includes a $14.3 billion investment in Scale AI, Alexandr Wang’s promotion as chief AI officer, Yann LeCun’s exit after 12 years, the reduction of 600 positions in FAIR research, and the establishment of Meta Superintelligence Labs, which launched its first closed-source model, Muse Spark, on 8 April.

      Meta has now secured five founding members from Thinking Machines Lab, the AI startup Mira Murati formed after her tenure at OpenAI. The latest exit, founding engineer Joshua Gross, joined Meta Superintelligence Labs in March after developing Tinker, the key API product of the startup. His departure came after co-founder Andrew Tulloch, who moved to Meta in October with a reportedly $1.5 billion compensation package over six years, making it potentially the most costly individual hire in tech history.

      The approach has been overt. Following Mark Zuckerberg's unsuccessful attempt to buy Thinking Machines Lab for about $1 billion, Meta began recruiting its founding members individually. Various sources have characterized this strategy as a "full-scale raid." It has proven effective, with five original members joining Meta, three returning to OpenAI, and one moving to Elon Musk’s xAI. Murati's venture, which raised $2 billion at a $12 billion valuation during a seed round led by Andreessen Horowitz in July 2025, was reportedly negotiating a new round at a $50 billion valuation by November, while losing a majority of its founding team.

      Notable exits and their new roles

      Tulloch and Gross are the most clearly documented departures regarding their new positions. Tulloch, an AI researcher who had previously worked for OpenAI, is now at Meta Superintelligence Labs, where he operates under Alexandr Wang, the former Scale AI CEO appointed as Meta's first chief AI officer in June 2025. Gross, with experience from OpenAI and Meta, now leads engineering teams in the same division.

      The remaining departures from Thinking Machines took varied paths. Barret Zoph and Luke Metz returned to OpenAI in January 2026, along with Sam Schoenholz. Zoph was reportedly dismissed by Murati for "unethical conduct" before immediately rejoining OpenAI. Devendra Chaplot transitioned to xAI in March. These exits have left Murati with a substantially restructured leadership team: she continues as CEO, Soumith Chintala, the creator of PyTorch who joined from Meta's FAIR lab, is chief technology officer, and John Schulman remains as chief scientist.

      Meta's revamped AI structure

      The talent acquisitions are part of a broader reorganization that has revamped Meta's AI division over the past year. In June 2025, Meta invested $14.3 billion for a 49% non-voting stake in Scale AI and appointed Wang to head the new Meta Superintelligence Labs division, alongside Nat Friedman, the former GitHub CEO. Zuckerberg referred to Wang as “the most impressive founder of his generation” in an internal memo.

      The restructuring has faced challenges. Yann LeCun, Meta’s chief AI scientist for 12 years and a prominent figure in deep learning, left in November 2025 after being asked to report to Wang. “You don’t tell a researcher what to do. You certainly don’t tell a researcher like me what to do,” LeCun told the Financial Times in January. He labeled Wang as “young and inexperienced” and cautioned that “a lot of people have left, and many who haven't will leave.” LeCun later raised $1 billion to establish AMI Labs in Paris, almost entirely sourcing the founding team from Meta’s AI research organization.

      By August 2025, Meta Superintelligence Labs was divided into four groups: TBD Lab for large language models led by Wang; FAIR for fundamental research; a division for products and applied research under Friedman; and an infrastructure unit led by Aparna Ramani. The AGI Foundations team, responsible for the Llama model family, was disbanded after the mediocre reception of Llama 4. LeCun publicly stated that the AI team had “fudged” some results. In October 2025, about 600 positions were eliminated from the FAIR and AI infrastructure units.

      The economics of the talent battle

      The compensation figures circulating in the AI talent market have escalated to a level that disrupts typical recruitment dynamics. OpenAI CEO Sam Altman has acknowledged that top researchers have been offered signing bonuses of up to $100 million. OpenAI’s chief scientist, Mark Chen, described Meta’s recruiting as “like someone