Beginning July 10, China’s Alibaba will prohibit staff from utilizing Anthropic’s programming tool Claude Code, as per various reports. 

Anthropic has already barred Chinese firms, along with foreign entities associated with those firms, from utilizing its models. The organization has reportedly been addressing loopholes that permit Chinese users access to Claude.

A recent Reddit thread indicated that some loophole mitigation involved a variant of Claude Code that could discreetly recognize Chinese users. Anthropic’s Thariq Shihipar mentioned in a post on X that this was “an experiment we launched in March aimed at preventing account misuse from unauthorized resellers and safeguarding against distillation.” (Distillation refers to a method where AI models are trained on the outputs of other models.)

“Our team has implemented more robust mitigations since that time, and we’ve actually intended to remove this for a while,” Shihipar stated.

Despite this, Alibaba has purportedly categorized Claude Code as high-risk software and is advising employees to utilize the company’s proprietary Qoder tool instead.