US, China launch US China AI Dialogue on incident alerts

The dialogue arrives after a summer of visible AI security failures and open calls for caution from the people building the technology. Bessent pointed to OpenAI's agents hacking Hugging Face as an example of frontier model capabilities and safeguards that have failed to keep pace, and he noted public safety concerns have grown since Trump's May visit to Beijing.

The political backdrop is split. Anthropic employees warned earlier this month that AI could be on course to wipe out humanity, and laboratory leaders including OpenAI's Sam Altman, Anthropic's Dario Amodei, and Elon Musk have called for a slowdown while safety standards are set. Amodei's essay asked for government help in international coordination, including with the Chinese Communist Party. Trump has mostly brushed off those calls, arguing his administration already has enough criminal and regulatory power over developers, and he has said a slowdown could cede the US lead to China. Nvidia's Jensen Huang said AI firms are already subject to cybersecurity and damage liability laws, and Hugging Face cofounder Thomas Wolf questioned whether the US can build global cooperation while wanting to widen its own lead.

What the mechanism can actually cover remains open. Washington is restricting exports of advanced chips and chipmaking technology to China, but Bessent said officials did not discuss those export controls in this weekend's meetings. Whether the dialogue moves from notification toward genuine alignment may hinge on whether the two sides can separate incident reporting from the broader technology rivalry, and the timing around Xi Jinping's visit to Washington this week is likely to shape how far the talks progress.