Google loses AI researchers to Anthropic and OpenAI
The departures of researchers linked to Gemini and AlphaFold show that the AI race is also a talent race. The critical question is not only who has the best models, but who can keep the people building them.
Summary
Google is losing high-profile AI researchers to direct rivals such as Anthropic and OpenAI, according to TechCrunch, citing a Bloomberg report.
Jonas Adler and Alexander Pritzel, both linked to Gemini development, are reportedly set to join Anthropic. The report follows Noam Shazeer's move to OpenAI and John Jumper's move to Anthropic. Jumper worked on AlphaFold and shared the 2024 Nobel Prize in Chemistry.
In practice
The AI race is no longer only about models, chips, and data centres. It is also a fight for small teams whose work can strongly affect model quality.
For Google, the risk is not being pushed out of AI. The risk is losing part of its research advantage as influential researchers move to smaller, more focused companies with stronger equity upside.
What we still don't know
There is no direct public confirmation from Google or Anthropic for all the reported moves. TechCrunch says it contacted Google for comment.
It is also too early to measure the practical impact on Gemini or Google DeepMind's research. Talent movement is normal in technology, but the profile of these departures and their destinations make this case more significant.
Why it matters
- It shows that AI advantage depends on specialised talent as much as infrastructure.
- It strengthens Anthropic and OpenAI as preferred destinations for top researchers.
- It suggests pre-IPO equity remains a powerful recruiting tool against large public companies.
- It puts pressure on Google to retain critical teams in a market where research has become product.