Starting July 10, 2026, Alibaba will prohibit its employees from using Anthropic’s programming tool, Claude Code. This decision aligns with Anthropic’s existing policy that restricts Chinese companies and their foreign subsidiaries from accessing its models. The company has been actively working to close loopholes that previously allowed Chinese users to utilize Claude.
In March, Anthropic implemented an experimental feature within Claude Code designed to identify Chinese users covertly. This measure aimed to prevent unauthorized resellers from abusing accounts and to protect against distillation—a practice where AI models are trained on outputs from other models. According to Anthropic’s Thariq Shihipar, the team has since developed more robust mitigations and had planned to remove the experimental feature.
In response to these developments, Alibaba has classified Claude Code as high-risk software. Employees are now instructed to transition to using Alibaba’s proprietary programming tool, Qoder, as a safer alternative.
This move underscores the ongoing complexities in the global AI landscape, where companies must navigate regulatory challenges and intellectual property concerns. By restricting the use of external AI tools like Claude Code, Alibaba aims to mitigate potential risks and assert greater control over its technological infrastructure. This decision also highlights the broader trend of companies developing in-house solutions to reduce reliance on third-party technologies, ensuring compliance with local regulations and safeguarding proprietary data.
