The leaders of the most prominent AI laboratories warned this week that uncontrolled agents could permanently diminish human agency over the future. Sam Altman of OpenAI and Dario Amodei of Anthropic addressed the UN Security Council on Wednesday, calling for the adoption of common global safety standards to coordinate oversight of frontier models. Amodei stated that AI poses a "risk to humanity as a whole" and pledged that Anthropic would slow the release of new technology to ensure safety, a commitment Altman also made on behalf of OpenAI.

The push for international coordination faced opposition from the U.S. government, with White House Office of Science and Technology Policy Director Michael Kratsios rejecting a global regulator as a "globalist scheme" and urging that development not be paused. Other figures, including researcher Yoshua Bengio and representatives from DeepSeek and Moonshot, also briefed the council on existential threats. Amodei noted that current technical trajectories could produce a "country of geniuses in the data center" within one or two years.