Even a Nobel Laureate: Anthropic Is Scooping Up Google DeepMind's Talent

Anthropic is hiring three core Google DeepMind researchers within just six days, according to Bloomberg and CNBC reporting in June 2026. John Jumper, the 2024 Nobel laureate in chemistry, and Jonas Adler and Alexander Pritzel, who helped build Gemini, are all heading to Anthropic. Over the same stretch, Transformer paper co-author Noam Shazeer left for OpenAI, making the drain of frontier AI talent from Google to its rivals unmistakable. This article lays out who went where, and what it means, strictly on the basis of primary reporting.

The Six-Day Timeline of an Exodus

Bloomberg reported on June 24, 2026, that Google is poised to lose two more Gemini staffers to Anthropic. The June 19 news of Nobel chemistry laureate John Jumper's move to Anthropic was followed on June 24 by the additions of Jonas Adler and Alexander Pritzel. Counting Shazeer's move to OpenAI, four senior Google researchers are leaving the company within six days. TechCrunch framed the trend as Google talent continuing to leave for rivals. Core members of the team that built Gemini are departing one after another.

The Weight of the Three Who Left

John Jumper is a Google DeepMind director who shared the 2024 Nobel Prize in Chemistry with Demis Hassabis for AlphaFold, the protein-structure-prediction model. A scientist who left a deep mark by predicting three-dimensional protein structures from amino acid sequences is now moving to Anthropic. Jonas Adler and Alexander Pritzel are researchers who played key roles in developing the Gemini model. Adler worked on Google's AI coding efforts and Pritzel handled pretraining, the early stage in which a model learns from large volumes of data, and both also contributed to AlphaFold research alongside Jumper.

The Pull of IPOs and Equity Incentives

The IPO preparations of OpenAI and Anthropic are cited as the backdrop for these moves. TechCrunch analyzed that the two companies are drawing talent with equity incentives ahead of going public. Recalling that Google spent roughly $2.7 billion in 2024 to bring Shazeer back from Character.AI, one can gauge how high the cost of securing talent has become.

How to Read This Signal

On the numbers alone it looks like a three-for-one swap, but what actually moved is not headcount so much as the texture of the resumes. That a Nobel laureate and a pretraining lead are heading to the same place suggests Anthropic is aiming at scientific applications and foundational model work at once. The fact that a pre-IPO company recruits with equity rather than cash matters too. Pre-listing stock is a bet on future value, so it reads as researchers wagering their own careers on that company's growth. The essence of model competition is once again confirmed to be talent competition, because where the people who built Gemini and AlphaFold go will help decide the direction of the next generation of models.

The Question It Leaves for Korea's Market

This pattern maps directly onto domestic AI organizations. In a structure where frontier labs absorb top-tier talent with equity and IPO upside, cash-salary-centric hiring in Korea struggles to compete for the same talent pool. Just as the move of a single Nobel-caliber researcher becomes news, this case shows that retaining and re-recruiting core talent is a problem to be solved through a company's growth narrative and equity design, not individual compensation alone.

Why We Resist a Premature Verdict

Bloomberg, CNBC, and TechCrunch are the primary outlets behind this reporting, yet the specific terms and start dates of the hires remain officially unconfirmed by the companies. How each researcher's role settles within a new organization will only become clear over time. For Google, the consecutive departures from the core Gemini team could be a burden. Still, it is too early to conclude that one or two moves translate into a technology gap, since organizational depth and follow-on hiring will determine the real impact. ASAP has organized this strictly on the basis of primary reporting, without speculating about terms that remain unconfirmed.

Sources: Bloomberg · CNBC · TechCrunch