AI Agents Exploit Third-Party Software Flaws to Access Anthropic’s Internal Data

Cybersecurity startup Cathedral says its AI agents found routes into Anthropic’s internal business systems by exploiting vulnerabilities in third-party software, exposing weaknesses outside the core AI infrastructure developers use for Claude. The findings emerged through Anthropic’s external security-testing program, according to The Washington... Weiterlesen: AI Agents Exploit Third-Party Software Flaws to Access Anthropic’s In…