US and China Eye AI Cooperation Amidst Global Dominance Race

Leaders of the United States and China, engaged in a competitive race for artificial intelligence dominance, are expected to meet this week. The summit occurs against a backdrop of increasing global apprehension regarding the potential threats posed by rogue AI.

Despite the intense competition, there are hopes for finding common ground on AI safety concerns. As China closes the technology gap with the U.S., both nations have voiced anxieties about the implications of AI in the hands of the other. "The U.S.-China conversation in AI is getting caught in the ups and downs in the broader U.S.-China relationship," noted Scott Singer, co-director of the China AI Initiative at the Carnegie Endowment for International Peace.

Ahead of the anticipated meeting between U.S. President Donald Trump and Chinese leader Xi Jinping, U.S. Treasury Secretary Scott Bessent met with Chinese Vice Premier He Lifeng. Following their discussions, Bessent indicated that the two countries had discussed establishing a mechanism for AI dialogue and proposed a new "notification mechanism" for AI incidents that could impact national security.

Potential discussion topics may include AI risks that could harm both nations, such as AI-enabled cyberattacks, biological misuse, major model failures, or loss of human control, according to Sun Chenghao, a senior fellow at the Center for International Security and Strategy at Tsinghua University in Beijing. Sun suggested that in these areas, "there is still meaningful room for cooperation" and that "it would be a mistake to dismiss the entire safety debate as geopolitical competition." He added that China is increasingly attentive to loss-of-control risks, malicious use, cyber threats, and the safety of AI agents.

Michael Kratsios, Trump’s former science and technology adviser, emphasized the importance of dialogue with China on AI, stating, "The key thing that we want to talk to them about is avoiding shared risks, and that’s something that I think we can have a fruitful conversation on."

Both governments share concerns about the potential use of AI to attack critical infrastructure. China's AI governance documents identify finance, electricity, telecommunications, and transportation as sectors vulnerable to frontier AI risks, according to Sun. Samm Sacks, a senior fellow at Johns Hopkins School of Advanced International Studies’ Institute for America, China, and the Future of Global Affairs, observed that the Chinese side has shown genuine interest in discussing issues like AI and cyber incident information sharing with U.S. experts. Xiao Qian, vice dean of the Institute for AI International Governance at Tsinghua University, believes that if the two countries can discuss reporting and communication protocols for AI-related emergencies, it "can help the two countries to build technical confidence over time."

Trump has previously stated, "whoever wins AI, wins," and stressed the need for the U.S. to maintain a technological edge over China, while downplaying calls to slow AI development due to safety concerns. Beijing, in turn, has accused the U.S. of seeking an AI monopoly and has restricted China's access to advanced chips. Experts suggest that any significant easing of U.S. export controls is unlikely, given each country's pursuit of competitive advantages. Alfredo Montufar-Helu of the consultancy Ankura noted that the room for U.S.-China collaboration on AI is "quite narrow."

Despite U.S. restrictions, Chinese AI models are increasingly challenging American frontier models. Companies like Moonshot AI and DeepSeek have developed models that, by some measures, rival those of U.S. counterparts such as Anthropic and OpenAI, often at a lower cost. These models are gaining global traction, including in the U.S. Professor Will Cong of Nanyang Technological University in Singapore commented that export controls have slowed China's expansion by limiting its compute capacity, but have not halted the development of competitive models.

China's domestic chip developers are reportedly catching up with the U.S., leading Chinese companies to rely more on locally produced chips. This development may embolden Xi in discussions with Trump, according to George Chen with The Asia Group.

Allegations of AI technology theft remain a point of contention. The U.S. government and AI companies have increased claims of "malicious" extraction of capabilities, or "distillation," from leading American AI models, assertions that Chinese officials reject. Anthropic has accused Chinese AI models from DeepSeek and Moonshot of routing user requests to its Claude AI models. Rebecca Arcesati, with the Mercator Institute for China Studies, suggests Chinese labs may be attempting to leverage the infrastructure of U.S. rivals due to U.S. restrictions on China's compute power. Chinese officials and some analysts argue that distillation is common practice and not the primary driver of Chinese models' advancements.

Beijing has also expressed concerns about threats to its critical infrastructure from American AI and the potential for AI to spread political misinformation and challenge its governance. Xiao Qian stated, "We cannot expect meaningful cooperation on AI safety if one side increasingly treats the other side’s technological progress as a national security threat." She added that disagreements between the U.S. and China "make dialogue even more necessary."