Alibaba Bans Claude Code

Alibaba has taken a significant step in its escalating conflict with Anthropic by banning the use of Claude Code, Anthropic's AI-powered coding agent, among its employees. The decision comes after security researchers discovered that Claude Code contained hidden tracking code designed to identify Chinese users. This code, which was present in the software since version 2.1.91 released on 2 April, checked system timezones and scanned proxy URLs against a list of Chinese domains and AI lab addresses.

The discovery of this tracking code has further strained the relationship between Alibaba and Anthropic, which was already tense due to allegations of industrial-scale distillation attacks by Alibaba on Anthropic's AI models. Anthropic had accused Alibaba of stealing its AI capabilities through these attacks, leading to a significant escalation in their conflict.

In response to the ban, Alibaba has recommended that its employees use Qoder, its own coding agent platform, as a substitute for Claude Code. This move not only reflects the company's efforts to distance itself from the controversial tracking code but also underscores its commitment to developing and promoting its own technologies.

The implications of this conflict extend beyond the immediate parties involved, as it touches on broader issues of AI security, intellectual property, and the ethics of data collection. As companies continue to develop and integrate AI technologies into their operations, incidents like these highlight the need for transparency, accountability, and robust security measures to protect user data and prevent unauthorized use.