China’s foreign ministry has confirmed that the US and China have agreed to establish official AI dialogue. The agreement came after direct, high-level communication between the two countries about AI regulation.
This high-stakes engagement resolves months of behind-the-scenes maneuvering. The US had been anxious about Anthropic’s frontier model, Mythos, demonstrating cyberweapon potential, while Chinese large language models gained leverage by adapting to Huawei chips, effectively sidestepping US compute restrictions.
Previously, a US lead negotiator publicly dismissed the idea of including other countries in setting global AI rules. This direct dialogue bypasses that tough stance and lays down a clear framework for crisis management — think Cold War-era hotlines, but for AI.
In its response, China emphasized that both major powers should work together to promote AI development and governance.