Anthropic Bans Chinese-Controlled Companies from Using Claude AI

Anthropic, the US artificial intelligence developer, announced on September 4 new AI service restrictions targeting Chinese-controlled firms. The updated terms bar any company with over 50% Chinese ownership—whether directly or via offshore subsidiaries—from accessing the Claude AI model. Anthropic cited legal, regulatory and national security risks, warning that data sharing requirements under Chinese law could expose its technology to military or intelligence use. The move underscores increasing global tension and divergence in AI regulation between the US and China. In response, China’s Zhipu AI quickly launched a “Claude API relocation plan,” offering seamless compatibility with its GLM model API and promotional packages to attract displaced Claude users. Industry analysts view Anthropic’s decision as part of a broader “soft-and-hard export control” strategy, following US chip embargoes and software export rules. While this shift may accelerate the development of domestic Chinese AI alternatives, it also highlights growing geopolitical fragmentation in the tech sector. Traders and investors should monitor the impact on AI stocks and cross-border data compliance risks as technology bifurcation accelerates.