Mira Murati Rejected $1 Billion from Meta—So Zuckerberg Reportedly Tried Poaching Her Team
Mira Murati rejected Meta’s $1B offer. Zuckerberg then reportedly tried poaching her team, offering co-founder Tulloch up to $1.5B.
After Mira Murati turned down Meta’s $1B offer, Mark Zuckerberg allegedly tried to recruit Thinking Machines Lab's top AI talent.
After Mira Murati turned down Meta’s $1 billion acquisition offer for her AI startup Thinking Machines Lab, Mark Zuckerberg didn’t take no for an answer. Instead, he allegedly launched a major recruitment effort to lure away the company’s top talent — beginning with co-founder Andrew Tulloch.
Meta's Aggressive Push After Murati’s Refusal
According to a report by The Wall Street Journal, following Murati’s rejection, Zuckerberg personally reached out to more than a dozen employees from Thinking Machines, a startup of just around 50 people. His primary target: Andrew Tulloch — a renowned AI researcher and co-founder of the company.
To win him over, Zuckerberg reportedly offered Tulloch a massive six-year compensation package worth up to $1.5 billion, tied to Meta’s stock performance and incentive bonuses. Despite the offer, Tulloch declined.
Meta has denied the scale of the package, with spokesperson Andy Stone calling the $1.5 billion figure “inaccurate and ridiculous.” He also emphasized that Meta “is not trying to buy Thinking Machines” and that any compensation would depend on stock value.
A Billion-Dollar Battle for AI Talent
Meta has been aggressively recruiting AI talent, not just from Thinking Machines, but also from other OpenAI spinouts like Anthropic, founded by Dario Amodei. According to WSJ, the company has approached over 100 OpenAI employees, successfully hiring at least 10.
Earlier reports from Wired revealed that Meta’s offers to Thinking Machines had ranged between $200 million and $1 billion, in hopes of bolstering its recently launched Superintelligence Lab.
Why Tulloch Matters in AI
Andrew Tulloch’s résumé explains the fierce bidding war. A University Medalist in mathematics from the University of Sydney, Tulloch holds a Master’s from Cambridge and a PhD from UC Berkeley. He previously worked at Meta (then Facebook) from 2012 to 2023, playing a key role in building core machine learning infrastructure, including co-developing PyTorch, one of AI’s most widely used tools.
After briefly joining OpenAI in 2023, where he contributed to GPT-4’s pretraining and reasoning systems, Tulloch partnered with Murati in early 2025 to launch Thinking Machines Lab — a startup that has now become the epicenter of Silicon Valley’s battle for top AI minds.
