OpenAI's and Anthropic's CEOs both told the UN that humanity could lose control of AI without global regulation

[NightCrawler]

Sam Altman of OpenAI and Dario Amodei of Anthropic both used platforms at the United Nations this week to call for global oversight of artificial intelligence, arguing that the industry's biggest decisions cannot be left purely to a handful of San Francisco based technology labs. Altman specifically warned that humanity could lose control of the future of AI, stressing that critical decisions must instead be shaped through democratic processes and by governments that are actually accountable to the people they serve, rather than by the companies building the technology itself.

Amodei struck an even starker tone, warning that AI, if mismanaged, could pose a risk to humanity as a whole, and underlining the urgent need for genuine international oversight rather than a patchwork of national rules or purely voluntary industry commitments. The pair were joined in spirit by Hugging Face co-founder Clement Delangue, who described using AI specifically to defend against AI driven attacks and noted that Chinese models currently face fewer restrictions than their American counterparts, and by AI researcher Yoshua Bengio, who sits on a UN scientific panel and warned that the dangers involved are real and imminent enough to demand genuine international cooperation. Foreign ministers from both France and the UK voiced support for building some kind of international control framework around the technology.

Not everyone at the UN shared that framing, however. Michael Kratsios, representing the Trump administration, rejected what he described as all efforts by international bodies to assert centralised control and global governance of AI, while China's own position remained focused squarely on technological competition rather than restriction, suggesting any genuine path toward binding global AI rules still faces serious geopolitical obstacles before it could ever become reality.

OpenAI, Anthropic CEOs call for global AI regulation at UNwww.aljazeera.com

[Ben55]

Having the CEOs of two of the biggest AI labs on the planet both publicly asking for outside regulation of their own industry is a genuinely unusual dynamic, most industries actively lobby against exactly this kind of oversight rather than requesting it.

[EventHorizon]

The Trump administration's rejection of any centralised global governance framework basically guarantees this stays a patchwork of competing national approaches rather than anything genuinely unified, at least for the foreseeable future.

[Ryan98]

Altman and Amodei calling for democratic accountability over their own industry's decisions is worth taking with at least some skepticism, it costs them very little to say this publicly while their companies keep shipping increasingly capable models regardless.

[Hannah56]

Chinese models facing fewer restrictions than American ones is a genuinely interesting asymmetry that rarely gets discussed openly, it complicates any argument that safety focused regulation alone will meaningfully slow down global AI capability progress.

[StringTheory88]

Bengio specifically using the words real and imminent is a notably stronger warning than the more hedged language usually used by researchers in this space, that phrasing alone is going to get quoted a lot

[Keira85]

China focusing purely on competition rather than restriction is the detail that actually undermines the whole push here, global regulation only works if every major player with real capability actually buys into it together