Top AI Researcher Andrew Tulloch Leaves Meta for Rival Anthropic
AI researcher Andrew Tulloch is leaving Meta for rival Anthropic less than a year after securing a massive compensation package, highlighting tech talent retention challenges.
In the fierce artificial intelligence talent race, billions of dollars are sometimes not enough to secure loyalty. Andrew Tulloch, one of the world's most sought-after AI researchers and a key developer of OpenAI's frontier models, is leaving Meta once again—less than a year after being re-hired in a staggering compensation package previously estimated at up to $1.5 billion. His new destination is rivaling firm Anthropic.
Tulloch's departure, coming immediately after Meta launched its personal AI agent Muse, highlights the paradox facing Mark Zuckerberg. Meta is projected to invest a massive $135–145 billion in computing infrastructure and chips this year, yet technological supremacy ultimately relies on a handful of elite scientists who understand complex training architectures.
In the 2026 AI market, top-tier researchers operate much like elite sports stars: moving fluidly between leading labs, accumulating immense bargaining power, and making career decisions based on research freedom and computing capabilities rather than bonuses alone.
Tulloch's exit adds to a series of turbulence within Meta's research divisions, underscoring that while computing power and data centers can be purchased with capital, the brightest minds are far harder to keep in-house.
