Google DeepMind’s Accelerating Talent Exodus to Anthropic

Gemini Co-Lead Noam Shazeer's Departure — to OpenAI, Not Anthropic

Noam Shazeer, co-author of the landmark "Attention Is All You Need" paper and co-lead of the Gemini models at Google DeepMind, announced on June 18, 2026, that he is leaving Google to join OpenAI — not Anthropic . This was Shazeer's second exit from Google; he had previously left to found Character.AI, and Google spent $2.7 billion to bring him back just two years ago . His departure to OpenAI, which is preparing for an IPO, underscores that the talent war is multi-front — Google is losing leaders to both Anthropic and OpenAI simultaneously .

Investor Fallout: Alphabet's Worst Trading Day in Over a Year

On Monday, June 22, 2026, Alphabet's stock fell approximately 5% — its worst single-day decline in over a year — wiping out roughly $225 billion in market capitalization . Bloomberg reported an intraday drop of up to 7.2% . CNBC noted that the slump was driven by "escalating worries regarding artificial intelligence and the recent departures of two prominent researchers" — referring to the back-to-back exits of Shazeer and Jumper . Morningstar reported that investors feared Google is "losing the war for AI talent" .

Broader Context: A One-Sided Talent Flow to Anthropic

A Fortune analysis from June 2025 found that DeepMind engineers were nearly 11 times more likely to leave for Anthropic than the reverse, while OpenAI engineers were 8 times more likely to defect to Anthropic . Other notable DeepMind-to-Anthropic moves include researcher Nicholas Carlini, who announced in 2025 that he was leaving DeepMind after seven years to join Anthropic for safety and security research .

Why Are They Leaving?

Multiple sources point to a combination of factors:

• Bureaucracy and slow decision-making at a large company like Google versus the agility of Anthropic
• Compensation packages — Anthropic has been aggressive with equity and salary offers
• Model competitiveness concerns — Google's Gemini models have lost ground to Anthropic's Claude and OpenAI's GPT on key leaderboards, raising doubts internally about Google's AI direction
• Mission alignment — several departing researchers cited Anthropic's safety-focused mission as a draw