Alibaba has reportedly circulated an internal notice instructing employees to stop using all Anthropic products and uninstall models like Claude Sonnet, Claude Opus, and Claude Code, along with the company's agent tools. The ban is expected to take effect on July 10.
Since the beginning of the year, Alibaba has been encouraging employees to use AI, offering free internal models and reimbursing external ones like Claude, GPT, and Gemini. Many developers had been mixing Claude Code and OpenAI Codex with Alibaba's own Qoder. This ban effectively removes Claude from Alibaba's internal development toolchain.
The shift comes amid Anthropic's increasingly aggressive efforts to restrict Chinese users. In June, Anthropic submitted materials to the US Senate accusing Alibaba of using a large number of accounts to interact with Claude at high frequency, labeling it an "industrial-grade model distillation attack." Since then, tensions have escalated, and many Chinese user accounts have been suspended.