Alibaba Internal Reverse Ban: Full Removal of Claude Series AI Tools

Alibaba has reportedly issued an internal directive banning all employees from using Anthropic's Claude series products, including models such as Sonnet, Opus, and Fable, as well as the intelligent agent tool Claude Code. The ban takes effect on July 10. Earlier this year, Alibaba was initially very open to AI applications. To accelerate its business intelligence transformation, the company not only provided free quotas for internal models but also offered generous reimbursement plans for external models. Under this policy, employees quickly increased their reliance on top-tier models like Claude and GPT, with some programmers consuming hundreds of dollars in call quotas per week. However, as risks around AI computing power and data security have become increasingly apparent, this open window has now been firmly closed. This "reverse ban" stems from a hidden conflict over AI technology security. Recently, the developer community disclosed a reverse analysis of Claude Code, revealing that since April, the tool has embedded a sophisticated detection mechanism. Its working logic involves injecting code to monitor users' device time zones and accessed domain lists in real time. If a user is detected in the China time zone or interacting with domestic tech companies' domains—including Alibaba—the system automatically reports the activity. Embedding such sensitive detection logic directly into the system prompt clearly crosses the red line of enterprise internal R&D security. In response, Anthropic officially stated that this mechanism was initially an "experimental" measure launched in March, aimed at preventing account abuse and model distillation attacks. The relevant code has been completely rolled back and removed in the latest version released on July 2.