The slow bleed of talent at Google has accelerated into a visible hemorrhage as of June 24, 2026, with four elite researchers departing within just three weeks. Jonas Adler and Alexander Pritzel, both viewed internally as key contributors to Google's Gemini AI model, are set to move to the Claude maker, according to people familiar with the matter. Adler drove high-stakes AI coding efforts while Pritzel anchored foundational model training. Within the last three weeks, Noam Shazeer, the Gemini co-lead who co-authored the seminal 2017 Transformer paper, defected to OpenAI after Google spent roughly $2.7 billion to acqui-hire him and his team from Character.ai. Days later, John Jumper, the 2024 Nobel co-winner for AlphaFold, ended a nine-year tenure at DeepMind to join Anthropic. The market is waking up to the fact that this represents not standard tech-sector churn; it is a targeted extraction of Google's brain trust.
The planned exits highlight the pressure Google faces from two startups that are on the cusp of going public, offering even well-heeled employees at Big Tech firms the chance at a rare payday by signing on before an IPO. Wall Street's reaction was immediate and violent, with Alphabet shares plunging between 5% and 7% across the June 22 trading sessions, wiping an estimated $180 billion to $270 billion from its market capitalization. At the current tape, GOOGL trades around $344.69, holding a $4.18 trillion market cap with a P/E of roughly 26.3x. The company had already lost two prominent staffers, with Nobel laureate John Jumper heading to Anthropic and star researcher Noam Shazeer going to OpenAI. Their moves rattled investors and cast new doubt on Google's ability to compete in the fierce race to build better models.
The talent exodus is intersecting dangerously with product execution issues, as Gemini 3.5 Pro, Google's flagship reasoning model promised for June, has now slipped to July. Management points to practical necessities: gathering tester feedback via Antigravity and LMSYS Arena, ironing out token-consumption bugs identified in the lighter 3.5 Flash release, and integrating new internal pretrain checkpoints (revision 25) cut off in March 2026. Under normal conditions, a one-month delay is a rounding error. But in the current climate, it calcifies a damaging narrative where OpenAI and Anthropic are dictating the cadence of the frontier, while Google appears bogged down in refinement and bureaucracy, playing defense.
The true driver behind the talent exodus is capital markets and structural convexity rather than purely cultural factors. Startups like Anthropic and OpenAI operate with a distinct financial weapon: pre-IPO equity arbitrage, offering senior researchers not only agility and focused mission, but a call option on becoming the defining platform of the decade. While Google can offer elite researchers exceptional cash compensation, liquid public stock, and massive compute clusters, it cannot offer the same structural upside potential. Anthropic recently raised a new round of funding at a $965 billion valuation, overtaking OpenAI, and is considering going public as soon as this fall, with intense secondary market activity and massive private valuations making top-tier researchers choose the "lottery ticket" upside of a highly capitalized neolab over the mature equity of a mega-cap incumbent.
A spokesperson for Google said the company remains confident in its position in the market for AI talent and pointed to Google DeepMind CEO Demis Hassabis's remarks earlier this week. 'There's a lot of movement between all the leading labs, and we attract our fair share of top talent. We have the largest and broadest research team in the industry. This is the most competitive talent market the tech industry has ever seen,' Hassabis said at an event in Cannes. Despite market concerns, analysts like D.A. Davidson's Gil Luria maintain that the actual risk is not that Google collapses, but that it stops being viewed as the frontier pace-setter. The company will survive and thrive, but its stock faces a real threat of multiple compression unless Gemini 3.5 Pro lands with undeniable force and management directly addresses the retention crisis.
