Google DeepMind co-founder Demis Hassabis is stepping down from his day-to-day role as CEO, transitioning to chair of the AI lab and chief scientist at parent company Alphabet, as the organization faces intense competition, heavy AI spending, and shifting commercial priorities.
The artificial intelligence landscape at Google is undergoing a massive structural overhaul. Demis Hassabis, who co-founded DeepMind in 2010 alongside Shane Legg and Mustafa Suleyman before Google acquired the company in 2014, is stepping back from his hands-on operational duties. Koray Kavukcuoglu, the lab’s chief technology officer, will replace Hassabis at the helm as senior vice president of Google DeepMind, reporting directly to Google CEO Sundar Pichai and taking over the day-to-day operations of the lab behind the company’s Gemini models.
Leadership Shift and the Push Toward Artificial General Intelligence
While handing over daily management, Hassabis is not leaving the company. Instead, he takes on the newly created role of chair of Google DeepMind and Alphabet’s chief scientist, while continuing to lead Isomorphic Labs, the tech giant’s AI-powered drug-discovery venture. In a blog post detailing the organizational changes, Hassabis explained his motivation for the transition, stating that he had been working toward AGI his whole life and now, like many others, felt it was close at hand. He noted that he decided now is the right time to hand over operational tasks to focus on the big picture and help shape what comes next.
Internal operations within DeepMind had already begun reflecting this division of labor. Kavukcuoglu had long been managing the technical development of Gemini, frequently leading discussions in company briefings while Hassabis attended only the most critical meetings. According to a DeepMind employee speaking to Time, Hassabis spent an increasing amount of time working on post-AGI readiness, safety, AI governance, and engaging with governments, including attending a recent G7 summit.
The company emphasized that safety remains central to the leadership transition. A Google spokesperson added that frontier model safety has lived directly within the Gemini team under Kavukcuoglu’s leadership from the very beginning, and that his teams will continue collaborating closely with safety and policy groups across Google and Google DeepMind.
Commercial Pressures, Spending, and Fierce Market Rivalry
The high-level reorganization arrives as Alphabet confronts questions regarding whether it can maintain its competitive edge against rival labs like OpenAI and Anthropic. Although Google’s cloud business has surged—with cloud revenue skyrocketing 82% year over year to $24.8 billion during the second quarter—capital expenditures are climbing steeply. Alphabet is forecasting between $195 billion and $205 billion in capital spending for the year, driven primarily by the infrastructure required to support its aggressive artificial intelligence ambitions.
.jpg?branch=production&width=3840&quality=75&auto=webp&crop=16:9 "Photo: time.com")
Market headwinds have also affected product timelines. Google delayed the release of its flagship model, Gemini 3.5 Pro, which had been expected to launch in June. Reporting from Bloomberg indicated the model fell short of specific internal goals, while Axios noted that low morale also contributed to the delay. Investors reacted nervously to the broader leadership shuffle, causing Alphabet’s share price to drop following the announcement.
In a December interview on the podcast The Tech Download, Hassabis described the competitive environment as ferocious
. That pressure highlights a core tension within Google’s strategy: Hassabis aims to use artificial intelligence to transform science and achieve artificial general intelligence, whereas Google must commercialize, monetize, and scale the technology fast enough to justify massive capital investments.
Scientific Triumphs and a High-Profile Departure
DeepMind’s scientific track record under Hassabis includes landmark milestones such as AlphaGo, which defeated the world’s best human player in the board game Go, and AlphaFold, an AI software capable of predicting protein structure folding critical for disease research and drug discovery. Hassabis noted that these breakthroughs relied heavily on Google’s vast computing power and proprietary chips.
Sebastian Mallaby, author of The Infinity Machine, noted that Hassabis’s shift back to strategic oversight isn’t out of character, pointing out that Hassabis historically fought for external oversight and prioritization of safety when selling DeepMind to Google.
Alongside the executive restructuring, the lab experienced a high-profile exit. Jeff Dean, a 27-year Google veteran who most recently served as Google DeepMind’s chief scientist, is leaving the company alongside several colleagues to launch a public-benefit corporation named Discovery Loop. Google plans to act as a founding investor and cloud partner for the new venture, which will focus on automating machine learning, science, and engineering.
Global Governance Proposals and What Comes Next
Beyond his corporate duties, Hassabis has advocated for systemic regulatory guardrails. In a blog post entitled A Framework for Frontier AI and the Dawning of a New Age, he argued that the world needs an international AI watchdog with the authority to evaluate frontier models before release and coordinate industry slowdowns if models pose severe risks. He proposed that the United States lead the initiative due to its technical and economic standing, suggesting a structure that could mirror regulators like the Financial Industry Regulatory Authority.
According to Axios, Hassabis has spent months quietly discussing the proposal with European officials, other AI labs, and the Trump administration, noting that he has received positive responses and hopes to establish the organization before the end of the year. As Kavukcuoglu steps into operational control of DeepMind, Google faces the dual imperative of accelerating product deployment and navigating the geopolitical and safety demands of an industry hurtling toward artificial general intelligence.
Related reading
