Sources: Google AI researchers Jonas Adler and Alexander Pritzel, both viewed internally as key contributors to Gemini, are planning to leave for Anthropic
Two leading artificial intelligence researchers at Alphabet Inc.'s Google are planning to leave for rival Anthropic PBC …
Context & Ripple Effects
This follows related coverage of John Jumper’s departure from Google’s AI coding work, which was described as adding strain to Google’s competition with Anthropic and OpenAI. The reported moves of Jonas Adler and Alexander Pritzel extend that personnel pressure beyond a single departure.
Related reporting also says Google has broadened an AI coding strike team into midtraining as it tries to catch up with Anthropic after executive departures. That makes the researchers’ planned exits relevant to both Gemini development and Google’s wider effort to close capability gaps.
First-order effects
- Google faces the prospective loss of two researchers regarded internally as important Gemini contributors, adding continuity and knowledge-transfer pressure to its AI teams.
- Anthropic would gain the planned hires from a direct rival, potentially strengthening its technical bench in areas where Google is seeking to improve.
Second-order effects
- Google’s expanded coding and midtraining effort may need to absorb more of the work and institutional knowledge previously held by departing researchers, increasing pressure on remaining teams and leadership.
- The moves sharpen the competitive value of experienced researchers who have worked on frontier-model programs, raising the stakes for retention and recruiting among leading AI labs.
Third-order effects
- If departures from incumbent AI organizations to specialized labs persist, competitive advantage will depend increasingly on the ability to assemble and retain concentrated research teams, not just on established platforms or brands.
- The pattern could reinforce a cycle in which talent movement and product catch-up efforts become mutually reinforcing: departures create execution gaps, while visible gaps make rival labs more attractive destinations.
The trend: This is one data point in the intensifying contest for frontier-AI talent, with specialized labs using recruiting to convert research momentum into a more durable competitive position.