OpenAI, Google DeepMind, Anthropic CEOs Converge on Mandatory Pre-Release Scrutiny and US-Led Global AI Governance Body

For the first time, the chief executives of OpenAI (Altman), Google DeepMind (Hassabis), and Anthropic (Amodei) converged in writing on the same regulatory framework — all three supporting mandatory external scrutiny of frontier AI models before public release and a US-led international AI governance body, per Axios reporting July 16. As recently as June the three companies occupied distinct regulatory positions; the convergence appears to be an attempt to shape a mandatory framework proactively, before Congress imposes one on its own terms. It lands as OpenAI’s 5% federal equity stake offer remains under negotiation, as the White House advances voluntary AI release standards with the labs, and as California has separately diverged from federal restrictions by contracting directly with Anthropic for state-agency access.

The structural question the convergence raises but does not answer is who actually staffs and controls the proposed “mandatory external scrutiny” body — an industry-authored governance framework, however binding-sounding, still leaves the composition and funding of the reviewing body to be determined by the same three labs now converging on the proposal.