Google DeepMind CEO Advocates for a U.S.-Led Regulatory Body for Frontier AI
Demis Hassabis, CEO and co-founder of Google DeepMind and Nobel laureate, has made a compelling case for the United States to establish a new independent standards body. This body would test and regulate the most potent AI models before their release. He warns that artificial general intelligence (AGI) is likely just “a few short years away.”
In a comprehensive post published on July 14, 2026, titled “A Framework for Frontier AI and the Dawning of a New Age,” Hassabis suggests that the new body should take after the Financial Industry Regulatory Authority (FINRA). FINRA is a self-regulatory organization that oversees U.S. broker-dealers, and it differs from a traditional government agency. The proposed Standards Body would primarily be funded by the tech industry. It would employ independent technical experts to test frontier AI models for dangerous capabilities. These include cybersecurity threats, biological risks, and the ability to bypass safety safeguards.
According to Hassabis’s framework, AI labs would initially voluntarily submit models up to 30 days before release for review. Once the testing protocol is proven robust, compliance could become mandatory for any frontier model seeking deployment in the U.S. market. The body could also coordinate an industry-wide slowdown if a model is deemed too risky. Hassabis hopes the body could be operational before the end of 2026.
The proposal comes as competitive pressure increases between U.S. and Chinese AI developers. It follows a White House executive order in June that already sought voluntary pre-release access to frontier models. Google DeepMind, Microsoft, and xAI have previously agreed to provide models for federal national security testing.
Source: TechCrunch – DeepMind CEO calls for an independent standards body to regulate frontier AI