Experts see scope for China-US cooperation on AI risks

China and the United States could use high-level exchanges to strengthen communication on artificial intelligence risks, taking practical steps to work together for the open and inclusive development of AI for all and for good.

Sun Chenghao, associate research fellow at Tsinghua University's Center for International Security and Strategy, said structural differences over trade, technology and security remain. But the two sides are increasingly exploring how to prevent friction in one area from disrupting the broader relationship — a central task in building a constructive China-US relationship of strategic stability.

In AI, Sun identified two strands in bilateral interactions: competition over chips, computing power, models, export controls, and shared risks, including AI-assisted cyberattacks, biological risks, and the risk of losing control over AI agents.

He said the high-level exchanges could provide an opportunity to establish practical arrangements for addressing common risks.

Those arrangements could identify responsible departments and establish regular channels of communication, clarify which incidents require notification, and provide for regular discussions between officials and technical experts. Given AI's reach across multiple sectors, such communication may require coordination among several departments, Sun said.

Zhou Bo, research fellow at the same center, said uncertainty about AI's potential harm makes dialogue particularly important. While its benefits are widely understood, its possible negative consequences are less clear, and that uncertainty can heighten fears.

"As long as China and the US are talking, that's a good thing," Zhou said.

As two leading countries in AI, China and the US have a responsibility to discuss the risks and, if possible, develop common rules, Zhou said. He said one possible area for progress would be a communication mechanism for AI-related crises, with clear conditions for when contact should be initiated.

Beyond risk management, Zhou highlighted AI's potential to benefit more countries, speaking highly of China's emphasis on open-source technology. He described China's international AI cooperation as largely focused on the development needs of the Global South, including open-source AI, infrastructure and capacity building. Such cooperation, he added, was also open to participation by developed countries.

If the high-level exchanges produce progress on AI, he said, that could provide a useful example for AI dialogue internationally.