Google DeepMind co-founder and CEO Demis Hassabis published a personal manifesto on July 14, 2026, titled “A Framework for Frontier AI and the Dawning of a New Age,” calling for the United States to establish an independent AI standards body modeled on FINRA — the Financial Industry Regulatory Authority, the private, industry-funded organization that polices Wall Street trading firms under SEC oversight. Hassabis’s proposal envisions a similar body for AI: an organization with the authority to test the world’s most advanced AI models before public release, screen them specifically for national-security risks, and — in the most consequential version of the proposal — coordinate an industry-wide slowdown if the risks from a specific model or capability level are judged serious enough.
Hassabis said such a body could realistically begin operating before the end of 2026 if the U.S. government moved to establish it, and that it should eventually have the authority to bar specific AI models judged unsafe from the American market entirely — a meaningfully more binding enforcement mechanism than the voluntary safety-testing frameworks other AI labs, including OpenAI, have separately been discussing with the White House this summer.
A Rare Moment of Public Agreement Among Rivals
What made the proposal notable beyond its substance was the reaction it drew. OpenAI CEO Sam Altman, whose company competes directly with Google DeepMind, publicly called Hassabis’s plan “thoughtful.” Elon Musk, who runs a competing AI effort through SpaceX’s xAI division and has sparred publicly with other AI leaders on regulatory questions, described it as “a good starting point for discussions” — a notable instance of two of the industry’s most prominent and often publicly disagreeing figures both offering genuine, if measured, endorsement of a rival’s regulatory proposal within days of its publication. Hassabis has said he intends to travel to Washington to meet directly with U.S. policymakers to advance the plan.
A Long-Term Vision Alongside the Near-Term Policy Push
Hassabis has paired his regulatory proposal with a genuinely optimistic long-term framing of AI’s potential: he has said publicly that he expects AI to usher in a new “golden era” of scientific discovery within 15 years, transforming medicine, energy, and space exploration. That optimism is reflected in concrete work already underway — Google DeepMind and Isomorphic Labs, the AI-driven drug discovery company Hassabis also leads, announced a joint approach this year specifically focused on “bioresilience”: using AI to help prevent bad actors from misusing biological research tools, while still letting legitimate scientists and governments benefit from the same underlying technology.
A Nobel Prize Sits Behind the Policy Credibility
Hassabis’s public standing on AI policy questions is reinforced by a genuinely rare scientific credential: he and DeepMind colleague John Jumper were jointly awarded the 2024 Nobel Prize in Chemistry for AlphaFold, the AI system that solved a roughly 50-year-old open problem in structural biology by accurately predicting how proteins fold into their three-dimensional shapes. That work now underpins Isomorphic Labs, the AI drug-discovery company Hassabis also leads, which closed a $2.1 billion Series B funding round in May 2026 led by Thrive Capital and has active partnerships with Eli Lilly and Novartis. Hassabis has said he expects Isomorphic Labs to have its first AI-designed drug candidates in human clinical trials before the end of 2026 — a concrete, dated commercial milestone for a company built directly on Nobel Prize-winning research.
What This Means for Philippine Founders
Hassabis’s FINRA-for-AI proposal is worth tracking closely for any Philippine AI startup building products on top of frontier models from any major lab: if a body with real authority to test and potentially block model releases in the U.S. market actually gets established this year, it would directly affect the pace and predictability of new model releases every Philippine AI product ultimately depends on, whether or not the Philippines has any equivalent regulatory body of its own. Isomorphic Labs’ own trajectory — from a Google DeepMind research spinout to a company with billions in outside funding and pharmaceutical partnerships within a few years — is also a concrete, credible example for Philippine biotech and health-tech founders of how far a genuinely novel AI application, built on real scientific breakthroughs rather than a general-purpose chatbot, can scale commercially, well beyond what a general-purpose chatbot product would ever be able to demonstrate on its own — a reminder that some of the most commercially significant AI applications may look nothing like the consumer products dominating most current AI headlines. The rare, near-unanimous praise from Altman and Musk is itself a signal worth taking seriously — when competitors who routinely disagree publicly on nearly everything else in AI policy independently converge on cautious support for the same proposal, it suggests a genuine, broad industry recognition that some form of binding, pre-release safety testing is coming, regardless of exactly which specific model it ends up being built around.
Share this article