U.S. Absence Stalls Global AI Governance Efforts at U.N.

International efforts to establish global guardrails for artificial intelligence hit a significant roadblock this week as the United States rejected participation in multilateral AI governance frameworks during the U.N. General Assembly.

President Donald Trump explicitly dismissed what he termed any "globalist scheme of control" on AI in his address to the assembly, creating an immediate challenge for the 22 nations—led by Finland and Norway—that adopted a declaration calling for international AI safety standards. Neither the U.S. nor China signed the declaration, despite hosting the vast majority of the world's leading AI laboratories.

The enforcement gap

The absence of American buy-in creates a fundamental problem for international AI regulation. Companies including OpenAI, Anthropic, and others developing frontier models operate primarily from U.S. soil, making any governance framework without American participation largely symbolic.

"If the U.S. is not participating — where the U.S. has a lot of the labs — and also is not focusing on this as a policy matter, it limits their effectiveness," Christopher Painter, former cybersecurity coordinator at the State Department, told POLITICO, which first reported these developments.

James Lewis, a former State Department official who worked on cybersecurity policy, put it more bluntly: "The people who make 90% of AI don't want to play."

Industry leaders sound alarm

The policy vacuum stands in sharp contrast to warnings from AI company executives themselves. At a U.N. Security Council session Wednesday, OpenAI CEO Sam Altman called for countries to "come together for shared interests and collective good in the face of a powerful new technology."

Anthropic CEO Dario Amodei went further, advocating for "common global standards" for AI model testing and warning that "if managed poorly, I believe that AI could be a risk to humanity as a whole."

Why it matters

The disconnect between industry concern and government action reveals a critical governance gap at precisely the moment when AI capabilities are advancing rapidly. Without coordination between the U.S. and China—where nearly all frontier AI development occurs—international safety standards become unenforceable recommendations. This leaves individual nations attempting to regulate a technology whose risks and benefits cross borders, while the countries with the greatest capacity to shape AI development remain outside the framework.

China's ambiguous position

Beijing has sent mixed signals. Fu Cong, China's permanent representative to the U.N., expressed support for "a consensus-based global governance framework," suggesting potential openness to international coordination. However, China did not sign the 22-nation declaration, and AI is expected to receive only passing mention during Chinese President Xi Jinping's visit to Washington this week.

U.N. Secretary-General António Guterres addressed the impasse directly, stating that "governments with the greatest AI capabilities have the greatest responsibilities to humanity."

The U.N. has established bodies including the Independent International Scientific Panel on AI, which will inform a Global Dialogue on AI Governance scheduled for May in New York. But creating binding international agreements remains a lengthy process—the U.N.'s cybercrime treaty took seven years from proposal to finalization.

European officials privately expressed frustration with the U.S. position, though few have criticized the Trump administration publicly. French President Emmanuel Macron called for nations to avoid becoming "vassals of either of the two great world powers" and proposed developing an open-source frontier model as an alternative.

These details were first reported by POLITICO.