Google DeepMind CEO、米主導のAI標準化機関設立を提唱

On July 14, Demis Hassabis, CEO of Google DeepMind, published an essay titled "A Framework for Frontier AI and the Dawning of a New Age" on his Substack, proposing the creation of a frontier AI standards body led by the United States. Hassabis characterizes the current moment as an "extremely important turning point in human history," arguing that AGI—a system possessing all cognitive abilities of the brain—will likely arrive within years. He forecasts that this era will be remembered as "the foot of the singularity," and that AGI's impact could be equivalent to "10 times the scale of the Industrial Revolution at 10 times the speed."

Hassabis identifies a central problem: fierce competition in the AI industry drives rapid progress, but the pace of frontier advances now outstrips the industry's and society's understanding of those advances. Cybersecurity challenges posed by frontier models are already visible, he warns, and nuclear and biological risks may soon emerge. Neither the industry nor society has secured the time and space needed for adequate risk mitigation. As a solution, he proposes establishing a "Frontier AI Standards Body" modeled on FINRA, the Financial Industry Regulatory Authority. This organization would be created as a public-private partnership or self-regulatory body under federal government oversight, with a board comprising independent technical experts and representatives from the open-source community. Funding would come primarily from industry, directed toward recruiting top technical talent and securing the computational resources needed for large-scale testing.

The standards body would develop evaluation protocols and work with federal agencies and US national laboratories to conduct testing in areas relevant to national security. Models meeting the body's benchmarks would be certified as "frontier-class," and organizations possessing them would be designated "frontier labs." Initially, frontier labs would voluntarily submit models for review up to 30 days before release; if the evaluation demonstrates efficacy, approval could become a requirement for deployment in the US market. In severe circumstances, the standards body could coordinate development slowdowns among frontier labs. Hassabis emphasizes that although he is British and based in London, he advocates for US leadership because "the United States is well-positioned, given its economic and technological standing, to take the first steps in developing such a framework." He frames a US-led initiative as a "powerful starting point" for building common standards globally, ultimately leading to international consensus. According to CNBC, Hassabis participated in a G7 meeting in June alongside Anthropic CEO Dario Amodei, where they called for a US-led coalition to shape AI rules and standards; Amodei has separately urged swift comprehensive policy action, including binding pre-release review.

Contextually, the US government issued export control directives on June 12 requiring Anthropic to halt access by foreign nationals to its latest models, "Claude Fable 5" and "Claude Mythos 5," citing national security authority; Anthropic suspended both models globally until the restrictions were lifted on June 30. Although Hassabis does not directly address this incident, his remark that "cybersecurity challenges posed by frontier models are already being seen" appears to reference it. Hassabis closes his essay with a call to action: "The future is not yet written. We must use this precious time before AGI arrives to shape this technology for the benefit of all humanity." The day before Hassabis published, over 200 economists and AI researchers, including 16 Nobel laureates, under the leadership of Stanford professor Erik Brynjolfsson, released a statement titled "We Must Act Now," warning that AI could trigger economic transformation far exceeding the Industrial Revolution in scale yet compressed into a much shorter timeframe, and urging swift construction of policy and institutions to ensure AI complements human capability and benefits society—though Hassabis did not sign this statement.

Hassabis's proposal reflects growing concern within the AI industry about the pace of development outstripping regulatory and risk-management capacity. The essay frames AGI as arriving imminently—within years—and compares its potential impact to a tenfold amplification of the Industrial Revolution in scale and speed. The CEO explicitly warns that frontier models already pose cybersecurity challenges, citing unspecified examples, and that nuclear and biological risks may soon emerge. By anchoring his proposal in an existing financial-sector regulatory model (FINRA), Hassabis suggests a pragmatic template rather than an untested framework.

The timing of the proposal is noteworthy: it follows Hassabis and Anthropic CEO Dario Amodei's participation in a G7 meeting in June (per CNBC), where they called for US-led coordination on AI rules and standards. Amodei has separately advocated for binding pre-release review in his own blog. The broader context includes the US government's June export controls on Anthropic's models (since lifted as of June 30), which Hassabis references obliquely when noting that frontier models' cybersecurity impacts are "already being seen." One day after Hassabis published his essay, over 200 economists and AI researchers, including 16 Nobel laureates, released a statement warning that AI could trigger economic transformation exceeding the Industrial Revolution in scale but arriving in far less time, calling for urgent policy and institutional action—though Hassabis did not sign that statement.