Google loses Adler and Pritzel to Anthropic as AI researcher exodus widens

Google is losing senior AI researchers to Anthropic at an accelerating pace. Jonas Adler and Alexander Pritzel, two prominent machine-learning scientists at Google, are departing for Anthropic, the Claude maker, according to reporting this week. The move extends a pattern of exits that has drawn away some of Google's most decorated AI talent over the past 18 months.

Adler and Pritzel join a growing roster of Google defectors now at Anthropic. Earlier departures include Noam Shazeer, who co-founded Character.AI after leaving Google, returned in an acqui-hire, and has since moved on again; and John Jumper, a Nobel Prize winner for his AlphaFold work, who joined a rival lab earlier this year.

What stands out

1. 01Anthropic is consolidating Google's research bench. The Claude maker has now recruited multiple former Google leads across protein folding, conversational AI, and generative modeling—suggesting a deliberate strategy to poach deep expertise rather than isolated departures.
2. 02The exits span research domains, not just one team. Departures across protein folding, conversational AI, and generative models indicate the problem is structural, not tied to a single project or manager.
3. 03Jumper's Nobel status made his exit symbolic. A laureate for foundational work on AlphaFold leaving Google signaled that even the company's most decorated scientists see better opportunities elsewhere.
4. 04Shazeer's path reflects AI talent fluidity. His cycle—Google exit to co-founder, acqui-hire return, and departure again—shows how quickly top researchers can move between roles and companies in a tight labor market.
5. 05