Hassabis Exit Sparks Power Shift at Google DeepMind Amid AI Capabilities Concerns

Google DeepMind's CEO Demis Hassabis stepped down from his role and took on a chairman position, citing a need to focus on long-term strategy. However, this move has been met with mixed reactions from employees, who are concerned about the lab's autonomy and direction.

Hassabis' departure comes at a critical time for Google DeepMind, which has been struggling to retain its technical edge in the AI race. The company has faced delays in releasing its Gemini 3.5 Pro model, missing three release deadlines in recent months. Engineers have blamed the delays on the company's failure to prioritize AI coding abilities.

The departure of Hassabis and several other top researchers has left employees wondering if Google has lost its way in the AI race. The lab is being absorbed into the rest of the company, with DeepMind's chief technology officer Koray Kavukcuoglu taking over day-to-day operations as senior vice president.

Google's models are currently not competitive with the bleeding edge of U.S. frontier on intelligence and coding benchmarks. Even some open-source Chinese efforts are outshining the tech giant. The company has been playing catch-up to rivals such as OpenAI and Anthropic, which have surpassed Google in AI capabilities.

A Google spokesperson disputed the characterization that power was shifting to Mountain View, saying Google DeepMind remained committed to its London presence. However, employees have expressed concerns about the lab's independence and autonomy under new leadership.