US artificial intelligence start-up Anthropic has updated its service restriction policy to prohibit access from global subsidiaries of Chinese companies – the latest limit imposed by a major player in an increasingly divided AI landscape.
Advertisement
In a statement published on its website on Friday, the San Francisco-based company laid out its updated terms of service, which now broaden access restrictions to entities “more than 50 per cent owned, directly or indirectly, by companies headquartered in unsupported regions”.
“This update prohibits companies or organisations whose ownership structures subject them to control from jurisdictions where our products are not permitted, like China, regardless of where they operate,” it said.
The changes mark an escalation from previous measures that targeted companies located within China.
Anthropic, the developer of the Claude AI models known for their coding capabilities, is not the first US AI start-up to restrict access from mainland China and Hong Kong. Last July, OpenAI also tightened its policies to block access from “unsupported countries and territories”.
Advertisement
Meanwhile, China has implemented a registration system for AI services. Currently, all approved models are domestically developed. The country has a vibrant open-source community for AI models, including those from DeepSeek and Alibaba Group Holding, owner of the South China Morning Post.