Amid growing international concern over the risks posed by artificial intelligence, the global conversation on A.I. governance is increasingly defined by tensions between the United States and China, leaving other nations largely marginalized in the debate. This dynamic has been highlighted during the U.N. General Assembly meetings in New York this week, where discussions on A.I. safety and regulation have occurred alongside a prominent state visit by Chinese President Xi Jinping to Washington.
While A.I. features prominently in U.N. dialogues focused on digital cooperation, economic implications, and governance, the absence of President Xi from the General Assembly to engage directly with President Trump—who has downplayed the severity of A.I. safety concerns—illustrates the heightened bilateral focus. The U.S. and China remain deeply divided over key issues such as intellectual property rights and technological restrictions, with the United States accusing Chinese firms of intellectual property theft and China protesting U.S. limits on advanced semiconductor exports essential for A.I. development.
Observers note that the race for A.I. supremacy mirrors Cold War-era tensions, with some experts drawing parallels to nuclear arms control negotiations. Foreign policy analysts characterize the current scenario as a “two-player game” dominated by the U.S. and China, leaving multilateral institutions like the United Nations sidelined from shaping meaningful international agreements.
In a recent development, the two powers initiated discussions on establishing a “U.S.-China A.I. dialogue,” a proposed notification system aimed at alerting one another to A.I.-related national security threats, signaling some degree of bilateral coordination despite broader rivalries.
United Nations Secretary-General António Guterres has underscored the existential challenges presented by A.I., alongside climate change and global inequality. He has emphasized the need for a shared international framework to ensure the safe and responsible advancement of the technology, highlighting risks such as misinformation, discrimination, privacy violations, and the potential for uncontrollable superintelligence. Despite his calls for collaboration, Guterres acknowledged that geopolitical divides hinder progress and that voluntary global pauses on A.I. development lack efficacy without the involvement of the U.S. and China.
Efforts at the U.N. have seen the endorsement of a declaration by leaders from 22 countries supporting enhanced human oversight of A.I. systems; however, both the United States and China abstained from this agreement. China's recent establishment of the World Artificial Intelligence Cooperation Organization in Shanghai reflects its ambition to assert greater influence over international A.I. policy discussions, contrasting with the United States’ relative disinterest in forging broader multilateral consensus.
The U.N. Security Council may play a more significant role in A.I. deliberations, with OpenAI CEO Sam Altman scheduled to address the council, which includes the U.S., China, and Russia. Nonetheless, experts express skepticism about the potential for consensus among these powers. Altman is also set to participate in a high-profile U.S.-China state dinner alongside prominent technology leaders such as Nvidia’s Jensen Huang, SpaceX’s Elon Musk, and Amazon’s Jeff Bezos.
Legal scholars note that the United Nations historically has faced challenges in influencing the behavior of dominant powers regarding weapon technologies, citing nuclear nonproliferation as an example. While the U.N. may struggle to directly shape U.S. and Chinese policies on A.I., it could succeed in ensuring that countries without advanced A.I. capabilities are not excluded from future technological benefits.
Ultimately, the evolving A.I. landscape reflects longstanding limitations within the U.N. system: while unable to compel action from the most powerful states, the organization remains a crucial forum for representing the interests of the wider international community.