Chinese tech giant Alibaba will reportedly ban employees from using Claude Code, Anthropic’s AI-powered programming tool, starting July 10, according to multiple reports.
The company has reportedly classified Claude Code as high-risk software and instructed employees to use Qoder, Alibaba’s own coding tool, instead.
Why Alibaba Is Taking This Step
The decision comes as Anthropic continues to tighten restrictions on access to its AI models by Chinese companies, as well as foreign entities owned by Chinese firms.
Anthropic already prohibits Chinese companies from using its models and has reportedly been working to close loopholes that allowed some users in China to access Claude indirectly.
The Controversy Around User Detection
The issue gained more attention after a recent Reddit post claimed that one version of Claude Code could secretly identify Chinese users.
In response, Thariq Shihipar from Anthropic said in a post on X that the feature was part of an experiment launched in March. According to him, the experiment was designed to prevent account abuse by unauthorized resellers and protect against distillation, a practice where AI models are trained using the outputs of other AI models.
Shihipar added that Anthropic has since introduced stronger safeguards and had already been planning to remove the experiment.
Alibaba’s reported ban reflects the growing tension around the use of Western AI tools inside Chinese companies, especially in sensitive areas such as software development and programming.
It also highlights a broader shift among major Chinese tech firms toward local AI tools, as access to U.S.-based AI models becomes increasingly restricted.
For Alibaba, directing employees to use Qoder is not only a technical replacement. It also appears to be part of a wider move to rely on internal tools that better align with the company’s security policies and operational requirements.
Read the article in
