Alibaba bans Anthropic Claude Code staff backdoor security July 10
Alibaba $BABA ordered staff to remove all Anthropic products by July 10 over alleged backdoor risks in Claude Code. Developers said the tool checked time zone and proxy settings to flag China-based users. Alibaba is steering employees to its in-house Qoder tool instead.
The rift between two major AI players just turned into a corporate ban. Alibaba $BABA has ordered employees to remove all Anthropic products, including Claude Code and the Claude model series, with the restriction taking effect July 10. The company cited alleged security risks after developers reported that Claude Code could inspect a user's computing environment, including time zone and proxy settings, to identify whether they were operating from China. An employee wrote on X that the mechanism began as a March experiment to curb unauthorized account resellers and model distillation, and would be removed in a future update. The ban deepens tensions that surfaced last month, when Anthropic accused operators tied to Alibaba's Qwen lab of running a large campaign to extract capabilities from Claude. Alibaba is now steering staff toward its in-house Qoder coding tool. The episode shows how AI supply chains are fracturing along geopolitical lines, with security audits becoming a lever in commercial competition.