A dispute over Anthropic's Claude Code coding tool has produced restrictions on both sides of the Pacific, according to reporting from oodaloop.com, the South China Morning Post, and The Decoder.
On the American side, Anthropic is trying to keep Chinese firms out. According to oodaloop.com, the company is moving to close loopholes that allow Chinese access to Claude. The Decoder reports that Anthropic is specifically attempting to block Chinese companies such as ByteDance and Ant Financial from using Claude Code — but that those firms are getting around the restrictions through VPNs and overseas subsidiaries.
On the Chinese side, a major tech giant is pushing the tool away on its own. According to the South China Morning Post, Alibaba has banned its staff from using Claude Code over what it describes as "spyware" concerns. The Decoder adds that Alibaba imposed the ban on its employees after hidden code was found in the tool.
The result is an unusual standoff: an American AI maker restricting who can use its product abroad, while a Chinese company restricts its own workers from touching the same product over data and security worries.
Why it matters: as AI coding assistants become embedded in how companies build software, they also become a new front in US–China technology tensions — where questions of data access, trust, and national origin can turn a popular developer tool into a geopolitical flashpoint.