Cybersecurity startup Cathedral reportedly used AI agents to infiltrate Anthropic's internal systems by exploiting vulnerabilities in third-party software during a bug bounty program, according to The Decoder.
Anthropic confirmed that the breach allowed access to sales, HR, and IT data but contested the extent of the intrusion, The Decoder reported. The use of AI in this context highlights evolving tactics in cybersecurity testing and threats.
For Japanese markets, where AI and cybersecurity are rapidly developing sectors, this incident underscores the growing importance of robust security measures around AI-driven tools and third-party integrations.