Alibaba Reportedly Bans Claude Code Over Alleged Backdoor Risk in AI Coding Tool

Reports allege that Anthropic's Claude Code included hidden logic to detect enterprise environments (e.g., proxy settings, timezones tied to Chinese corporate networks) and subtly encode detection results via prompt/punctuation changes; Alibaba has reportedly moved to ban the tool internally while Anthropic developers describe the feature as an anti-abuse control to be removed, and no independent audit has yet confirmed malicious intent or active exploitation.