“We Could Lose Control”: AI Leaders Deliver a Stark Warning to the UN

What Are AI Leaders Asking Governments To Do?

Amodei called for several specific international measures. These include an agreement preventing AI from being used to develop biological weapons, systems for evaluating and verifying model capabilities, and common standards for testing loss-of-control and misuse risks. He also called for a notification system for major AI-related security incidents.

Altman proposed complementary national and international frontier-AI standards. These would measure AI capabilities, assess risks, determine whether safeguards are adequate and preserve meaningful human oversight as systems become more autonomous. He also called for common evidence standards and secure channels for governments to share emerging vulnerabilities.

Hugging Face CEO Clément Delangue also addressed the council remotely. He argued that open-source AI can help defenders respond to AI-enabled attacks and highlighted the importance of reducing the gap between attackers and defenders.

AI governance and responsible adoption are becoming strategic priorities for technology leaders. The Chief Technology Officer & AI Leadership Programme from upGrad can help technology professionals develop the leadership, AI strategy, and decision-making skills needed to navigate the rapidly evolving AI landscape. 

The U.S. Pushes Back on Global AI Controls

The calls for international safeguards did not receive unanimous support.

White House Science Adviser Michael Kratsios said concerns about AI control were not a reason to pause development or create a new global governance structure. He argued that the U.S. wants allies and partners to share in AI's benefits.

China's UN ambassador Fu Cong also called for countries to have freedom in choosing AI technologies and promoted open-source models, while warning against forcing countries into an AI technology divide.

The meeting therefore exposed a growing divide: AI companies are asking governments for stronger safeguards, while some governments remain concerned that international controls could restrict technological development and national competitiveness.

For now, the UN discussion has produced calls for cooperation and standards, rather than a binding global AI agreement. The debate over who controls increasingly capable AI  governments, companies or both,  is still unresolved.