Woofun AI reports that Alibaba has instituted a total ban on all Anthropic Claude series products, with enforcement commencing on July 10. Employees are directed to remove Agent applications, including Sonnet, Opus, Fable, and Claude Code, marking a reversal of previous policies that subsidized external model usage.

This directive follows mid-June allegations from Anthropic regarding "industrial-grade model distillation attacks" via fake accounts, which led to account suspensions for Chinese users.

Additionally, recent disclosures revealed that earlier Claude Code versions contained covert mechanisms collecting data from Chinese tech domains, a practice Anthropic later acknowledged and rolled back.