The United States is reportedly preparing to curb American companies' use of Chinese-made artificial intelligence models, according to CNBC. An unnamed State Department spokesperson said such use "raises serious concerns," arguing that Chinese AI is "designed to advance Beijing's narratives, censor dissent and reflect CCP ideology and values." The move matters because a growing number of US firms are turning to Chinese models, drawn by lower costs and quality comparable to domestic offerings, raising questions about how far Washington can go in restricting private-sector choices.

The report cited Coinbase chief executive Brian Armstrong, who noted the firm's use of GLM 5.2 (from Z.Ai) and Kimi 2.7 (from Moonshot), while a separate CNBC report quoted the head of startup Lindy, which switched to DeepSeek to control costs. Nikkei Asia reported a surge of American adopters, including Airbnb and Uber, coinciding with Anthropic's government-prompted suspension of two models, Mythos 5 and Claude Fable 5. It remains unclear whether the US could impose a sweeping ban beyond changing its own procurement rules, given potential First Amendment issues around open-source models and the difficulty of policing overseas operations. Tensions run both ways: Reuters reported that China has held similar talks to limit exports of homegrown AI, and Beijing's Ministry of Industry and Information Technology has alleged that Anthropic's Claude Code contains a backdoor posing a "serious threat."