Skip to main content

Alibaba Bans Claude AI Tools Amid Security Concerns

Alibaba has issued an internal directive banning the use of Anthropic's Claude series AI tools, including models like Sonnet, Opus, and Fable, as well as the Claude Code agent. Employees must comply by July 10.

Earlier this year, Alibaba encouraged the use of external AI models to boost productivity, offering free quotas for internal models and reimbursement for external ones. This led to heavy reliance on tools like Claude and GPT, with some programmers spending hundreds of dollars weekly on API calls. However, growing concerns over AI computing power and data security have prompted a strict reversal.

The ban stems from a hidden detection mechanism discovered in Claude Code. Since April, the tool reportedly monitored users' device time zones and accessed domain lists, flagging those in China or interacting with domestic tech domains like Alibaba. This violated internal security policies.

Anthropic stated the mechanism was an experimental measure to prevent account abuse and model distillation, and it has been removed in the latest version released July 2.

Image

Key Points

  • Alibaba bans all Claude AI tools by July 10.
  • Hidden detection in Claude Code raised security alarms.
  • Anthropic says the feature was experimental and has been removed.