OpenAI and Anthropic CEOs press UN Security Council for global AI guardrails

The heads of two leading artificial intelligence companies told the UN Security Council that the technology's growing power demands stronger oversight, urging governments to build guardrails before frontier systems outpace the institutions meant to govern them.

OpenAI chief executive Sam Altman and Anthropic chief executive Dario Amodei addressed the Council in New York, both arguing that the benefits of advanced AI should not come at the cost of losing human control over it.

Altman described the present moment as one of both extraordinary promise and understandable anxiety. He said AI models could help uncover new knowledge, advance medicine, improve healthcare and education, and make economies more productive, while also serving as a tool against the growing complexity of modern life.

At the same time, he warned that as such systems become more capable and more autonomous, they could move faster than institutions, concentrate power in too few hands, or take decisions that people no longer understand or control.

He outlined two ways AI progress could go badly wrong, the first being a loss of control over the future if the technology advances too quickly for people to follow or intervene. The industry, he said, must not accept excessive technological risk simply because the benefits appear too important to slow down, and competing in a race is no reason for rash choices.

Altman said his company has unilaterally slowed down in the past and would do so again, adding that developers should avoid building systems whose alignment, monitorability and safety cannot be guaranteed. He stressed that no single person, company or country should be able to use the most powerful models to impose a worldview on others, and that people must remain at the centre of AI decision-making.

Amodei called for international cooperation and proposed a step-by-step approach. He suggested starting with narrow agreements that every member state could support, including a ban on using AI to develop biological weapons. He also urged the creation of evaluation and verification systems that keep pace with AI development, so that states can see frontier model capabilities and check each other's commitments.

He further proposed common global standards for testing models against loss-of-control and misuse risks, along with a notification system for AI incidents with significance for global security. Describing the issue as the most important global security challenge today, he said differences among nations must be set aside to confront both the opportunity and the threat.

Neither executive endorsed an extinction-level scenario, but both made clear that the power of AI systems should be harnessed under regulation rather than allowed to develop unchecked.