Google DeepMind’s Hassabis calls for US-led global AI watchdog

Demis Hassabis, co-founder and CEO of Google Alphabet Inc. ↗ 0.08% DeepMind, is calling on the United States to establish a new global AI regulatory body with the power to screen the world's most advanced AI models and pause their development if they are deemed dangerous, according to a report published Monday.yahoo+2

The proposal represents Hassabis's most concrete regulatory prescription to date, moving beyond his previous warnings about AI risk toward a specific institutional framework. It comes as international debate over how to govern frontier AI systems has intensified in recent months.

Building on G7 Momentum

Hassabis has been laying the groundwork for this push throughout 2026. At a Stanford University talk in May, he described AI as a "species-level transition" that leaves humanity with "little margin for error" over the next decade. He argued for increasing international coordination on AI regulation within five to ten years.stanforddaily+1

At the G7 Summit in Evian-les-Bains, France, last month, Hassabis and Anthropic CEO Dario Amodei jointly advocated for a US-led coalition to shape global AI rules during a closed-door meeting that included President Donald Trump. G7 leaders subsequently committed to enhanced collaboration on frontier AI risks, assigning finance officials, regulators, and cybersecurity specialists to evaluate how advanced AI models could affect financial stability and labor markets.firstpost+2

A Crowded Governance Landscape

The proposal enters an increasingly active regulatory environment. In June, the Trump administration issued an executive order titled "Promoting Advanced Artificial Intelligence Innovation and Security," directing agencies to develop a classified benchmarking process to assess advanced AI capabilities and setting thresholds for designating "covered frontier models". The order relies on voluntary cooperation between AI developers and the federal government rather than mandatory licensing.mcdermottlaw+2

At the international level, the United Nations launched its Independent International Scientific Panel on AI in February 2026, comprising 40 members tasked with producing evidence-based assessments of AI risks and opportunities. The panel presented its first annual report at the inaugural Global Dialogue on AI Governance in Geneva earlier this month.un+1

Whether Hassabis's proposed watchdog would complement or compete with these existing frameworks remains an open question, as does how such a body would reconcile US leadership with the multilateral approaches favored by European and developing nations.