- Alibaba bans employees from using Anthropic’s Claude Code.
- Ban follows concerns over Claude Code inspecting user environments.
- Anthropic accused Alibaba of attempting AI model distillation.
Alibaba has reportedly asked its employees to stop using Anthropic’s Claude Code and switch to its in-house coding assistant, Qoder, after internally classifying the tool as “high-risk software.” The ban is expected to take effect from July 10. The move adds to the already strained relationship between the two companies, which have been at odds over AI model training practices.
Neither Alibaba nor Anthropic has issued a public statement on the reported workplace restriction.
Why Did Alibaba Flag Claude Code As High-Risk?
According to Reuters, the decision follows concerns raised by developers over certain built-in mechanisms in Claude Code that appeared to inspect user environments, including time zones and proxy-related details, before sending prompts to Anthropic’s servers. This raised alarms among users in China, where Anthropic officially restricts access to its services.
ALSO READ: WhatsApp’s New Green Dot On iPhone Tells You Exactly Who Is Online Right Now
Responding to those concerns on X, Anthropic engineer Thariq Shihipar clarified that the feature was “an experiment we launched in March,” aimed at preventing account abuse by unauthorised resellers and protecting its models from distillation.
What Is The Dispute Between Anthropic And Alibaba About?
The reported ban also comes against the backdrop of a broader conflict. Last month, Anthropic accused Alibaba of attempting to extract capabilities from its AI models through distillation, a process where a smaller model is trained on outputs from a more advanced one.
In a letter to two US senators, reviewed by Reuters, Anthropic alleged such efforts could accelerate Chinese AI companies’ access to capabilities found in its advanced models.
ALSO READ: Amazon Prime Day 2026: 5 Deals Worth Checking Before The Sale Closes Tonight
Separately, Microsoft is also reportedly scaling back Claude Code use internally, though for different reasons, primarily to consolidate tools and cut costs by shifting developers to GitHub Copilot CLI. Unlike Alibaba’s case, this is not driven by security concerns. Anthropic’s models will still be available through that platform.
