Altman Floats Merging Top AI Labs Into One Body for Safety

The proposal is notable less for its feasibility than for what it concedes: that frontier-risk oversight is being discussed as a structural question of corporate control, not just a technical or policy one. The timing undercuts it — with OpenAI and Anthropic both preparing IPOs, competitive incentives are hardening precisely as coordination would need to loosen them. The parties most affected are the labs themselves and the safety teams inside them, whose influence depends on where such a body would sit. Whether the idea gains any traction beyond an internal float, or quietly fades as listing pressures mount, is the open question.