- Published on
Altman, Musk, and Hassabis back Amodei's call to add independent oversight
- Authors
- Name
- DeepHorus
On September 13, 2026, OpenAI CEO Sam Altman, Elon Musk, and former DeepMind CEO Demis Hassabis publicly backed Anthropic CEO Dario Amodei's call for independent oversight inside AI laboratories. The rare cross-industry endorsement signals growing consensus that external evaluation of AI systems is necessary. Amodei has long argued that labs cannot be trusted to police themselves, and his proposal has now gained support from some of the most influential figures in the field.
Musk's support comes as little surprise, given his years of public warnings about AI risks. More notable is Altman's endorsement, as OpenAI has often positioned itself as a leader in safety research. Altman and Amodei agree that AI labs should be evaluated independently, a stance that could reshape how frontier models are tested and deployed.
In a related development, Altman told Fortune that OpenAI will not go public this year, citing safety concerns. He had already shared that decision internally in June. However, another reading suggests OpenAI's financials simply do not support an IPO right now, especially compared to Anthropic's performance. Both explanations could be true, reflecting the complex pressures facing leading AI companies.
The push for independent oversight has implications for the entire industry. If adopted, external evaluators could set standards for model releases, slowing down the race to deploy ever-larger systems. The IPO delay also highlights the tension between commercial growth and safety commitments.
Not everyone agrees with the proposed approach. Google researcher Peyman Milanfar pushed back, arguing that recursive self-improvement is a naive assumption because feedback loops are inherently unstable. He contends that the harder a system optimizes itself, the deeper it falls into blind spots, and that reliable evidence comes only from the real world, not benchmarks. Milanfar called Amodei's "speed limit" unnecessary, stating: "Systems that reliably improve themselves will be governed, damped, slow, and bounded by margins that look wasteful. Because stability is the speed limit."
The debate between external oversight and natural stability is likely to define AI governance in the coming years. While the backing from Altman, Musk, and Hassabis gives Amodei's proposal significant momentum, skeptics like Milanfar argue that technical realities may already provide the necessary constraints. The industry now faces the challenge of translating consensus into concrete policy without stifling innovation.
