ai
Alibaba Bans Employee Use of Anthropic AI Tools From July 10
The Chinese e-commerce company will require staff to uninstall Anthropic models and switch to its internal Qoder assistant. The policy follows accusations from Anthropic of a large-scale distillation effort targeting its Claude system.
airedale.futurecdn.netAudio version
Tap play to generate a narrated version.
Developing·Limited corroboration so far. This page will refresh as more sources emerge.
Alibaba will ban employees from using Anthropic artificial intelligence tools for work purposes starting July 10. CNBC reported that the company placed Anthropic's Claude Code on a high-risk software list after an internal evaluation identified security vulnerabilities.
Staff must uninstall all Anthropic models and agent products, including the Sonnet, Opus, and Fable families, according to people familiar with the matter.
Employees are instead directed to use Alibaba's own AI assistant, Qoder. The decision follows a June letter Anthropic sent to the U.S. Senate Committee on Banking, Housing, and Urban Affairs. In that letter, Anthropic accused Alibaba of carrying out the largest known distillation attack to date and of attempting to extract its AI capabilities through nearly 25,000 fraudulent accounts.
Anthropic's terms of service already prohibit Chinese companies and other adversarial nations from accessing its models. Both companies declined to comment on the ban. The Financial Times reported Friday that Anthropic is closing loopholes that allowed Chinese firms to reach Claude through third countries.
Chinese fintech group Ant had provided employees corporate Claude accounts accessed via its Singapore entity, while ByteDance began reimbursing engineers for personal subscriptions on April 2 to encourage exposure to a wider range of AI products. ByteDance engineers access the service through virtual private networks. Ant and ByteDance declined to comment on those reports.
