US and China reportedly weigh AI risk talks ahead of leaders’ summit

The United States and China are considering launching official discussions on AI risk management, The Wall Street Journal reported, citing people familiar with the matter.

According to the report, the White House and the Chinese government are also considering whether to place AI on the agenda for a planned summit in Beijing between US President Donald Trump and Chinese President Xi Jinping. If agreed, the talks would mark the first AI-specific engagement between the two governments under the current US administration.

The possible dialogue could focus on risks linked to advanced AI systems, including unexpected model behaviour, autonomous military applications and misuse by non-state actors using powerful open-source tools, people familiar with the discussions told the newspaper. The report said Washington is waiting for Beijing to designate a counterpart for the talks.

The WSJ reported that US Treasury Secretary Scott Bessent is leading the US side, while Chinese Vice Finance Minister Liao Min has been involved in discussions on setting up such a channel. The newspaper added that the two presidents would ultimately decide whether AI appears on the formal summit agenda.

Liu Pengyu, spokesperson for the Chinese Embassy in Washington, was cited as saying that China is ready to engage in communication on AI risk mitigation. Analysts have raised the possibility that any future dialogue could support crisis-management tools, including an AI hotline between senior leaders.

The report places the latest deliberations in the context of earlier US-China engagement on AI. In 2023, then US President Joe Biden and Xi launched a formal AI dialogue, and both sides later said humans, not AI, would retain authority over nuclear-launch decisions. The WSJ said the earlier process produced limited results, but AI has remained a high-level focus in bilateral relations.

Non-governmental discussions have also reportedly continued in parallel, including exchanges involving former Microsoft research executive Craig Mundie and Chinese counterparts from Tsinghua University and major AI companies. Participants cited by the newspaper said those exchanges have focused on frontier-model safety, technical guardrails and broader questions of strategic stability.

Why does it matter?

A formal AI risk channel between Washington and Beijing would signal that both governments see advanced AI as a strategic stability issue, not only an economic or technological race. Even brief talks could matter if they create channels for crisis communication about military AI, frontier-model failures, or misuse by non-state actors. However, because the discussions are still only reported as under consideration, the significance lies in the possibility of a risk-management mechanism, not in any confirmed diplomatic breakthrough.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!