Andrew Tulloch, Meta’s Star AI Recruit, Is Leaving After Less Than a Year
Andrew Tulloch, the researcher whose reported $1.5bn offer from Meta became a symbol of the AI talent war, is leaving the company after less than a year, according to Semafor. The Wall Street Journal confirms he is joining Anthropic, though neither company has officially announced the move.
September 10, 2026 – 5:42 pm
Credit: gguy / Shutterstock.com
Tulloch’s departure comes just a year after he joined Meta, with his reported pay offer making him a prominent figure in the AI talent landscape. Reed Albergotti from Semafor broke the news on Wednesday, citing a source with knowledge of the situation. The Wall Street Journal followed up on Thursday with details of his move to Anthropic.
Personal Website Still Lists Meta Affiliation
Interestingly, Tulloch’s personal website still lists him as part of Meta’s TBD Lab, working on "superintelligence," as of Thursday.
Background and Return to Meta
Tulloch has a strong background at Meta, having spent 11 years there before moving to OpenAI in 2023. His website states that he contributed to training GPT-4o, GPT-4.5, and o3 while at Meta. He later co-founded Thinking Machines Lab with Mira Murati, OpenAI’s former CTO.
Contested Offer Value
Meta has consistently denied the $1.5bn offer figure, with spokesperson Andy Stone calling it "inaccurate and ridiculous." Tulloch initially rejected the offer but joined Meta months later. The exact sum he accepted remains undisclosed.
Implications for Meta and Anthropic
If the Journal’s report is accurate, this timing is awkward for Meta, which has been heavily investing in its AI team. They recently brought in five founding members from Thinking Machines Lab. Tulloch’s departure raises questions about the longevity of talent acquired through financial means alone.
For Anthropic, it represents a significant hire following a week that also saw the departure of researcher Jacob Coxon, who raised concerns about the risks of AI labs.
Wider Churn in AI Labs
The move is part of a broader trend of talent churn in AI labs. Google chief scientist Jeff Dean left last month to start a new company, and several senior researchers have followed suit. The limited pool of experts who have trained frontier AI models means each lab transfer carries valuable knowledge to competitors.