Alibaba will bar employees from using Anthropic's AI tools for work as of July 10, citing what it calls back-door security risks and putting Claude Code on a high-risk software list, according to CNBC. Anthropic's June letter to a US Senate committee had accused Alibaba of a ‘distillation attack' to illicitly extract its AI capabilities, an allegation Alibaba has not publicly addressed. Chinese companies were already barred from Anthropic's models under its terms of service, and Alibaba is now directing staff to uninstall Anthropic products in favor of its own Qoder assistant. Online criticism in China has grown over unverified claims that Claude Code included hidden code to detect whether users were based in the country, which Anthropic has not confirmed.
