China Sets Preconditions for AI Safety Talks With U.S.

China Demands Regulatory Parity Before AI Dialogue

China has outlined stringent preconditions for artificial intelligence safety discussions with the United States, demanding Washington demonstrate that American AI companies face the same regulatory scrutiny as their Chinese counterparts before substantive talks can proceed.

The conditions were published Sunday on Yuyuantantian, a social media account affiliated with Chinese state broadcaster CCTV that frequently signals official policy positions. The post insisted the U.S. must prove its AI firms are subject to equivalent safety, disclosure, and audit requirements.

"A clear distinction must be drawn between genuine security threats and mere technological competition," the post stated, adding that this line "must be drawn jointly by all participating parties."

The statement comes as officials from both nations prepare for AI-focused talks ahead of Chinese President Xi Jinping's planned state visit to the U.S. on September 24 for a summit with President Donald Trump. No specific date has been announced for the AI dialogue, and Chinese officials were still seeking agenda clarity last month to determine appropriate representation, according to details first reported by Bloomberg.

Anthropic Singled Out for Criticism

The Chinese government post specifically targeted Anthropic, accusing the San Francisco-based AI company of overstepping user data boundaries, conducting covert monitoring, and transmitting website domains without authorization. The post, titled "Anthropic Has Contracted the American Disease," argued that America's frontier AI models "have already developed in a distorted direction."

Chinese officials have grown increasingly concerned about Anthropic's Mythos model, viewing it as a potential offensive weapon while questioning why China is denied access for ordinary purposes, Bloomberg previously reported.

Anthropic blocked its services from Chinese-controlled companies last year, citing U.S. national security concerns. The company has also accused Alibaba of illicitly accessing Claude through thousands of fraudulent accounts—an allegation the Chinese tech giant has not addressed. Neither Anthropic nor the U.S. State Department immediately responded to requests for comment.

Why It Matters

This public positioning reveals how deeply intertwined technical AI safety discussions have become with broader U.S.-China strategic competition. By framing the talks as requiring regulatory reciprocity rather than accepting American-defined standards, China is challenging U.S. efforts to set global AI governance norms. For companies operating in both markets, this signals that AI safety frameworks will remain fragmented along geopolitical lines rather than converging toward universal standards.

Broader Containment Concerns

The Yuyuantantian post characterized U.S. AI policy as part of a broader effort to contain China's technological advancement. "The 'control' proposed by the US is, in essence, an attempt to make China accept an order partly defined by American companies," it stated.

The post also criticized the U.S. government for "repeatedly retreating" from establishing proper security boundaries around AI development, suggesting Washington has failed to regulate its own industry while seeking to impose restrictions on China.

These details were first reported by Bloomberg and the Los Angeles Times.