Alibaba reportedly bans employees from using Claude Code
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Alibaba will ban employees from using Anthropic's Claude Code programming tool starting July 10, according to multiple reports.
- Anthropic already prohibits Chinese companies and foreign entities owned by them from using its models, and has reportedly been working to close loopholes that let Chinese users access Claude.
- Anthropic ran a version of Claude Code in March that could secretly identify Chinese users; spokesperson Thariq Shihipar called it "an experiment" meant to prevent unauthorized-reseller abuse and protect against model distillation.
- Shihipar said the team has since implemented "stronger mitigations" and had been meaning to take the covert detection version down for a while.
- Alibaba has classified Claude Code as high-risk software and is instructing employees to use the company's own Qoder tool instead.
Why it matters: One of China's largest tech companies is moving thousands of engineers off a leading US AI coding tool onto a domestic alternative (Qoder) by July 10 — a concrete, dated instance of US–China AI decoupling that goes beyond rhetoric. Anthropic, meanwhile, has been actively engineering against Chinese access, including a covert detection experiment it has since moved to retire.
