Back to all articles
•Technology

Alibaba Bans Anthropic's Claude Code Over Security Concerns

View original source

China's Alibaba Group has announced that it will prohibit the use of Anthropic's Claude Code by its employees starting July 10 due to security concerns, specifically the presence of potential 'backdoors'.

  • Anthropic's Restrictions: Anthropic has already barred Chinese companies and foreign entities owned by them from using its models. This move is part of their effort to enforce stricter measures against unauthorized access.
  • Backdoor Allegations: A recent Reddit post accused Claude Code of embedding spyware that could identify Chinese users. Thariq Shihipar, a representative of Anthropic, clarified that this was an experiment implemented in March to prevent account abuse and unauthorized reselling.
  • New Tools: In response to the security threats, Alibaba has opted to utilize its in-house programming tool, Qoder.
  • Risk Classification: Alibaba has categorized Claude Code as high-risk software, which prompted the strict internal control measures.