China warns of "security backdoor" in Anthropic AI coding tool

Chinese cybersecurity officials have identified a potential vulnerability in Anthropic’s Claude AI coding assistant, characterizing the flaw as a risk that could allow unauthorized access to sensitive software environments. The concern centers on how human-in-the-loop AI systems process external code, highlighting broader industry challenges regarding the security of AI-integrated development tools. While no specific exploits have been confirmed, the warning underscores rising international scrutiny over the integration of generative AI into critical infrastructure and software supply chains.