Altman calls for US-led global AI safety body in FT op-ed

OpenAI CEO Sam Altman published an opinion piece in the Financial Times on July 1 titled "This is how we can make AI safe for everyone," proposing a US-led international framework to establish global AI safety standards. The piece argues that determining safety standards is "a prerequisite to broad distribution" of advanced AI systems, and calls on elected representatives — not AI labs — to set the rules governing the technology.threads+2

A Post-G7 Push

The op-ed arrives two weeks after Altman joined fellow AI executives Dario Amodei of Anthropic and Demis Hassabis of Google Alphabet Inc. ↗ 0.78% DeepMind at the G7 summit in Évian-les-Bains, France, where the trio sat alongside heads of state from the world's largest democratic economies. At that gathering, Altman warned leaders, "Do not cede your responsibilities to AI labs like mine," while Hassabis called specifically for "a U.S.-led standards body that ideally works with close cooperation with the international democratic community".brookings+2

The concept echoes a body Altman has previously compared to the International Atomic Energy Agency — an organization that would coordinate safety testing and governance norms among democratic nations while potentially engaging rival powers including China. The White House has simultaneously accelerated its own plans for AI model standards, according to the Financial Times.kucoin+1

Competition and Governance Converge

The governance push comes as the three leading AI labs are locked in an intensifying competition across model capability and enterprise adoption. OpenAI is also preparing for an initial public offering, adding financial incentives to seek regulatory clarity. Anthropic and Google DeepMind have broadly endorsed the same international coordination framework, though Amodei has argued that collaboration should encompass structured access to frontier models and chip trade while explicitly excluding China from certain arrangements.stocktwits+2

The Enforcement Question

Policy analysts have noted that calling for standards is only the first step. A Brookings Institution analysis published June 30 argued that any resulting framework must be enforceable — not voluntary — drawing parallels to the Financial Action Task Force and Basel III banking standards, both of which emerged from prior G7 action and carried real consequences for noncompliance. The analysis urged governments to expand any standards process beyond technical concerns to include "behavioral standards" addressing competition, innovation, and the public interest, with civil society participation alongside industry.brookings