Google is bleeding AI talent to Anthropic, and it’s getting worse. Top researchers Jonas Adler and Alexander Pritzel are the latest to jump ship, joining a growing exodus that includes AI luminaries Noam Shazeer and John Jumper. The departures signal mounting competitive pressure in the race to build the next generation of AI models, with smaller, well-funded rivals poaching the very scientists who helped establish Google’s early dominance in machine learning.

Google just lost two more star researchers to Anthropic, and the pattern is becoming impossible to ignore. Jonas Adler and Alexander Pritzel are heading to the AI safety startup, joining what’s turning into a significant brain drain from the company that essentially invented the transformer architecture powering today’s AI revolution.

The timing couldn’t be more awkward for Google. These departures come on the heels of exits by Noam Shazeer and John Jumper, two scientists whose work fundamentally shaped modern AI research. Shazeer co-authored the landmark “Attention Is All You Need” paper that introduced transformers, while Jumper led the AlphaFold project that cracked protein folding. Losing researchers of this caliber isn’t just about headcount – it’s about losing the institutional knowledge and creative firepower that drives breakthroughs.

Anthropic, founded by former OpenAI executives including siblings Dario and Daniela Amodei, has positioned itself as the thinking person’s AI lab. The company’s focus on AI safety and constitutional AI resonates with researchers who worry about the breakneck pace of development at larger competitors. But make no mistake – Anthropic is also competing directly with Google in the enterprise market, where its Claude models challenge both OpenAI’s GPT-4 and Google’s Gemini.

The talent war in AI has reached fever pitch, and Google is discovering that being first doesn’t guarantee you’ll stay on top. Microsoft has OpenAI, Amazon is pouring billions into Anthropic, and even Apple is quietly assembling its own AI dream team. Google’s once-unassailable position as the place to do cutting-edge machine learning research is under siege.

What’s driving researchers out the door? Industry insiders point to a mix of factors. Google’s bureaucracy has grown alongside its ambitions, making it harder to move fast and take risks. The company’s cautious approach to AI deployment – born partly from genuine safety concerns and partly from regulatory anxiety – contrasts sharply with the move-fast ethos at startups like Anthropic. Compensation matters too. While Google pays well, equity packages at pre-IPO AI startups offer potentially life-changing upside.

Adler and Pritzel aren’t household names outside AI circles, but their work carries weight. Both have published influential research and contributed to core projects within Google’s DeepMind division. Their departure suggests the exodus isn’t limited to the absolute biggest names – it’s spreading across Google’s research ranks.

The competitive landscape has shifted dramatically since Google’s researchers published that transformer paper in 2017. Back then, Google looked unstoppable in AI. Today, OpenAI dominates mindshare with ChatGPT, Anthropic is winning enterprise contracts with its safety-first pitch, and even Meta is making waves with its open-source Llama models. Google went from defining the field to playing defense.

Anthropic’s hiring spree comes as the company raised $7 billion in its latest funding round, with backing from Amazon, Google itself (ironically), and other major investors. That war chest gives Anthropic the resources to make offers Google researchers can’t ignore. The company is building out its team ahead of what sources describe as major product launches planned for later this year.

For Google, the challenge isn’t just retention – it’s perception. Every high-profile departure feeds a narrative that the company has lost its innovative edge in AI. That perception makes the next departure more likely, creating a talent spiral that’s hard to reverse. Google still employs thousands of world-class AI researchers, but momentum matters in Silicon Valley. Right now, the momentum is flowing toward smaller, hungrier competitors.

The brain drain also raises strategic questions. Can Google maintain its technical edge if its best people keep leaving? Does the company need to restructure its AI efforts to move faster and empower researchers? Some inside Google argue the company should spin out its AI division or give it more autonomy, similar to what it did with Alphabet’s Other Bets. But that kind of corporate surgery is easier said than done.

What happens next will shape the AI industry’s power dynamics for years. If Anthropic successfully absorbs this talent and ships breakthrough products, it validates the startup model over big tech incumbency. If Google manages to stem the bleeding and ship compelling AI products despite the exits, it shows that institutional resources still trump individual star power. Either way, researchers like Adler and Pritzel are voting with their feet, and right now they’re walking toward Anthropic’s door.

The steady stream of researchers leaving Google for Anthropic tells a bigger story than just personnel changes. It reflects a fundamental shift in where AI’s best minds want to build the future. Google created the technology powering the AI boom, but it’s struggling to keep the people who made that possible. For Anthropic, every hire from Google isn’t just about adding talent – it’s about establishing legitimacy and momentum in a market where perception often becomes reality. The question isn’t whether Google can survive these departures – it’s whether the company can adapt fast enough to stop the next wave from walking out.