Anthropic
Citing a sharply worded analysis, Bloomberg reported that Yuyuantantian, a prominent social media channel affiliated with state broadcaster CCTV and often utilised to signal Beijing’s strategic policy stance, insisted that Washington must verify its domestic AI enterprises comply with equivalent safety, disclosure and independent auditing standards before any meaningful cross-border discussions can occur. Chinese state media has leveled sharp accusations against American artificial intelligence (AI) company Anthropic, accusing it of unethical practices while setting rigid prerequisites for upcoming bilateral AI negotiations with the US.
The post said: “A clear distinction must be drawn between genuine security threats and mere technological competition. In a piece titled “Anthropic Has Contracted the American Disease,” the publication argued that leading US models have evolved down a distorted path, with Washington repeatedly failing to uphold baseline safety guardrails. The state-backed commentary argued that both countries must collectively establish the boundary between legitimate security hazards and ordinary technological rivalry, the report added. This line must be drawn jointly by all participating parties. Yuyuantantian directly singled out Anthropic and its flagship AI model, Claude, alleging that the AI model breaches user privacy limits, conducts covert user monitoring and transmits external web domain records without permission. The dispute highlights deepening unease among Chinese officials regarding advanced Western AI models. They have expressed apprehension that Anthropic’s Mythos system can be deployed offensively against national infrastructure, questioning why commercial access to the platform remains blocked for standard purposes in China.
Anthropic restricted its platforms from Chinese-controlled organisations last year, pointing to US national security priorities.