Google DeepMind CEO Calls AI a Species-Level Risk, Urges Global Rules

What Happened

Demis Hassabis, co-founder and chief executive of Google DeepMind, publicly warned this week that artificial intelligence represents a "species-level" risk to humanity and called on governments to establish binding global rules to govern its development. The statement, reported across multiple outlets, marks one of the most direct public warnings issued by a sitting executive at one of the world's leading AI research organizations.

Background

Hassabis co-founded DeepMind in London in 2010. Google acquired the company in 2014 for a reported 400 million dollars. DeepMind has since become one of the most prominent AI research laboratories in the world, responsible for systems including AlphaFold, which predicted the three-dimensional structure of nearly every known protein, and AlphaGo, which defeated world champion Go players. Hassabis shared the 2024 Nobel Prize in Chemistry for the AlphaFold work.

Google DeepMind operates under Alphabet, which reported first-quarter 2026 revenue of 90.2 billion dollars, with its Google Cloud division, which houses much of its AI infrastructure, posting 12 billion dollars in revenue for the same period.

The call for international AI governance is not new among researchers and policymakers, but it carries added weight when made by the chief executive of a laboratory that is itself actively developing frontier AI systems. Hassabis has previously expressed concern about AI safety in academic and conference settings, though his remarks this week were characterized as among his most urgent public statements on the subject.

What He Said

Hassabis described AI as a "species-level" concern, a framing that places the technology alongside other existential risks that have historically prompted multilateral treaty frameworks, such as nuclear weapons and climate change. He urged that international rules be established before AI capabilities advance further, though specific treaty structures or regulatory bodies were not named in the available wire report.

Hassabis did not specify which governments he was addressing or outline a timeline for the governance proposals he described. No draft framework or formal proposal was attached to his remarks as reported.

The Broader Context

Hassabis's warning arrives during an active period of AI policy debate. The European Union's AI Act entered its phased enforcement period in 2024 and 2025, establishing the first comprehensive statutory framework for AI in a major economy. In the United States, executive and legislative efforts to regulate frontier AI models have proceeded more slowly, with no comprehensive federal AI law enacted as of mid-2026.

Separately, the United States and China recently opened bilateral talks specifically addressing AI risks, reflecting growing governmental recognition that advanced AI development crosses national boundaries in ways that may require coordinated responses. International bodies including the United Nations and the Organisation for Economic Co-operation and Development have each issued AI governance recommendations, though none carry binding enforcement authority.

Other senior AI executives have made similar public statements in recent years. OpenAI CEO Sam Altman testified before the United States Senate in 2023 and called for government oversight of AI. Anthropic, founded by former OpenAI researchers, has structured its corporate charter around what it describes as the responsible development of AI for long-term human benefit.

What Happens Next

Google DeepMind is expected to present further research findings at upcoming scientific conferences later in 2026, and international bodies including the United Nations are scheduled to convene additional working sessions on AI governance frameworks in the second half of the year.