China Says US AI Dialogue Impossible Without Accountability and Transparency

China has said that it will set strict rules before a big meeting between the two countries. China declares that talks about intelligence (AI) with the US cannot happen until Washington shows that its own AI makers follow the same rules, for safety, checking and openness. This strong statement was posted on a media page that is linked to the state broadcaster CCTV, which often shows Beijing’s official view. Chinese commentators also criticised the behaviour of Anthropic, a US AI developer and they set out the rules that should guide future diplomatic talks.

China sets strict conditions ahead of diplomatic talks

Representatives from Beijing and Washington are expected to hold specific talks on AI prior to Chinese President Xi Jinping’s state visit to the US on September 24 for a scheduled summit with US President Donald Trump. Although the exact date has not yet been announced, Chinese authorities requested further details regarding the planned agenda last month before finalising the composition of their delegation. According to the report, the growing tension between the world’s two largest economies was evident in the state media analysis, which specifically singled out Claude, Anthropic’s flagship system.

The publication accused Anthropic’s AI models of violating user privacy boundaries, conducting covert surveillance, and transferring web domains without authorisation. In an article titled “Anthropic Has Contracted the American Disease,” the commentary argued that cutting-edge US models have taken a distorted path and accused the US government of repeatedly evading its responsibility to enforce genuine safety limits. The text stated that bilateral negotiations cannot be limited to mere technical discussions, as Washington seeks to impose its own corporate safeguards as universal standards, using technological controls to curb China’s digital advancement and enforce compliance with rules drafted by US companies.