Full Breakdown
Google Faces Talent Exodus as Top AI Researchers Join Anthropic and OpenAI
6/25/2026, 8:43:48 PM
Key Departures from Google’s Gemini Team
In late June 2026 Google confirmed that Jonas Adler and Alexander Pritzel, senior researchers who led Gemini’s code-generation and pre-training work, will leave for Anthropic. Their exits follow Noam Shazeer’s move to OpenAI and DeepMind director John Jumper’s shift to Anthropic. All four were identified as pivotal contributors to Google’s large-language-model program.
Recent Talent Flow and Underlying Drivers
Google’s Gemini model, a large-language-model effort, has become a central focus of the company’s AI work. Since early 2026 a series of high-profile exits has occurred, with analysts citing pre-IPO equity offers from Anthropic and OpenAI as a primary lure. Internal reallocations of tensor-processing-unit (TPU) resources, such as the diversion of compute from Shazeer’s project to a DeepMind team, have also been reported as contributing factors.
Timeline of Exits
- Early June 2026: Shazeer announces move to OpenAI.
- Mid-June 2026: Jumper confirms departure for Anthropic.
- 24 June 2026: Bloomberg reports Adler and Pritzel will join Anthropic.
- 25 June 2026: Market reaction and analyst commentary reported.
Market and Financial Impact
Alphabet’s share price slipped 0.30 % after the Adler-Pritzel announcement. Jumper’s exit was reported to have erased roughly 6 % of Alphabet’s market value—about $245 billion—in a single session. Bloomberg and other outlets note that pre-IPO equity in Anthropic and OpenAI promises returns that exceed the predictable gains from Google’s restricted stock units.
Official Statements
A Google spokesperson referenced DeepMind CEO Demis Hassabis, who affirmed the company’s continued ability to retain AI talent. Anthropic declined to comment on the hires. Google has not released further details about the departures.
Analyst Concerns
Analysts warn that the talent drain could blunt Google’s AI momentum, especially as teams vie for limited TPU capacity. One analyst linked the compute reallocation from Shazeer’s project to a London-based DeepMind group with his decision to leave. Critics also highlight a broader “talent war” where pre-IPO equity at smaller labs appears more attractive than the stability of a $4 trillion public firm.
Implications for Google and the AI Landscape
The exits raise questions about Google’s capacity to maintain its competitive edge in large-language-model development and related scientific work. Market participants have responded with heightened volatility, reflecting concerns that loss of core researchers could affect Google’s ability to develop new models and its competitive positioning against Anthropic and OpenAI.
Future Outlook
Anthropic and OpenAI are expected to file for IPOs between late 2026 and 2027, potentially intensifying recruitment. Google’s leadership has signaled plans to strengthen internal talent pipelines and review compute-allocation policies. Observers will track subsequent hiring trends and any further impact on Alphabet’s stock performance.
