U.S. artificial intelligence company Anthropic has expanded its restrictions on entities linked to the Chinese mainland, banning access to its services for subsidiaries and overseas-based organizations majority-owned by companies in "unsupported regions." The move, announced September 5, cites "legal, regulatory, and security risks," and now applies to entities with over 50% ownership by firms in China, Russia, North Korea, or Iran, regardless of their operational locations.
Anthropic, valued at $183 billion and backed by Amazon, stated that some groups circumvented prior restrictions by using foreign-registered subsidiaries. Nicholas Cook, an AI-focused lawyer with experience in China, noted this marks the first formal public prohibition by a major U.S. AI firm but predicted limited immediate commercial impact due to existing market barriers and local AI adoption.
Chinese Foreign Ministry spokesperson Guo Jiakun responded by opposing the "politicization of trade and sci-tech issues," emphasizing such measures "do no one good." Meanwhile, Anthropic’s executive revealed the policy could cost the company "low hundreds of millions" in revenue.
The San Francisco-based firm, known for its Claude chatbot, recently settled a $1.5 billion copyright lawsuit over AI training data. Analysts suggest the restrictions reflect growing U.S.-China tech tensions, with implications for global AI development and cross-border tech collaboration.
