Google DeepMind CEO Demis Hassabis has called for the U.S. to establish a standards body to oversee advanced AI models and assess risks, including cybersecurity and biological threats. In a post on X, Hassabis emphasized the need for 'urgent action' to address challenges posed by artificial general intelligence (AGI), which he believes could be just a few years away.
Hassabis proposed a U.S.-led public-private partnership to review frontier AI models before public release, citing the country's economic and technical standing as key advantages. He suggested the body could resemble the Financial Industry Regulatory Authority (FINRA), a private, nonprofit Wall Street watchdog, and likely be industry-funded.
The call follows recent interventions by the Trump administration to limit the release of AI models by companies like OpenAI and Anthropic. Hassabis has long advocated for expert oversight of AI advancements, warning that rapid progress requires a dynamic, adaptable, and rigorous approach to testing.
Deeper Dive & Context
Proposed Framework and Stakeholders
Hassabis' proposal comes after discussions with tech leaders and heads of state, including at a G7 meeting where he and other AI executives called for a U.S.-led coalition to shape AI rules. The White House, State Department, and Department of Commerce have been approached for comment.
Industry and Government Responses
The Trump administration has already taken steps to regulate AI model releases, signaling a more hands-on approach to governance. Meanwhile, other AI leaders, such as OpenAI's Sam Altman, have also called for similar regulatory bodies, though specifics vary.
Long-Term Implications
Hassabis has previously predicted AGI could arrive by 2030, stressing the need for interdisciplinary collaboration to address broader societal impacts. He has gathered philosophers, economists, and other experts to consider the long-term consequences of AI advancements.
Diverging Perspectives
While Hassabis advocates for a U.S.-led body, some argue that global cooperation is essential to ensure equitable and effective AI governance. Others question whether industry-funded oversight could lead to conflicts of interest.