Industry

AI-generated text

Demis Hassabis steps back from day-to-day DeepMind leadership to focus on AGI research

Alphabet announced that Demis Hassabis will relinquish daily operational control of Google DeepMind to concentrate on long-term AGI research as Google DeepMind President and Alphabet’s lead scientist.

Demis Hassabis steps back from day-to-day DeepMind leadership to focus on AGI research

Alphabet has announced a leadership shift at Google DeepMind: Demis Hassabis will hand over daily operational responsibilities as CEO of Google DeepMind.

Hassabis’s new roles and responsibilities

Demis Hassabis will serve as Google DeepMind President and as Alphabet’s lead scientist. He will also continue to oversee Isomorphic Labs, the company’s unit focused on drug discovery. According to the official explanation, Hassabis is stepping back because he believes humanity is approaching the development of artificial general intelligence (AGI) and wants to concentrate on the “big picture” rather than day-to-day management.

Successor and departing figures

Koray Kavukcuoglu, currently Chief Technology Officer of Google DeepMind, will succeed Hassabis as CEO. Separately, Jeff Dean — who worked at Google for 27 years and recently served as a DeepMind lead scientist — is leaving the company with several colleagues to launch a public-benefit organization called Discovery Loop.

Background and milestones

Demis Hassabis co-founded DeepMind in 2010 with Shane Legg and Mustafa Suleyman; Google acquired the company in 2014. In 2024 Hassabis and two colleagues received the Nobel Prize in Chemistry.

Commercial pressures, finances and product delay

The leadership change comes amid criticism that Google has fallen behind competitors such as OpenAI and Anthropic in AI development. At the same time, Google’s cloud business reported strong revenue: second-quarter cloud revenue rose 82 percent year over year to $24.8 billion.

Alphabet also expects large capital expenditures this year, projecting between $195 billion and $205 billion, primarily for AI-supporting infrastructure. Additionally, Google postponed the planned June launch of the flagship Gemini 3.5 Pro model; Bloomberg reported the delay was due to the model failing to meet the performance levels leaders expected.

Why this matters

The organizational shifts highlight a transition in DeepMind’s and Google’s AI strategy: leadership reorganization, a delayed flagship model, criticism about competitive positioning, and significant planned investments together shape how the company’s AI efforts may evolve in the coming years.

(Synthesized from company announcements and contemporary reporting.)