Cathedral says AI agents found routes into Anthropic systems within 12 hours

First reported by Cyber Security News at · Updated · 2 sources

The startup says its automated agents identified ways into Anthropic’s internal business systems. The work was an authorized test under Anthropic’s program for outside researchers, according to Cyber Security News. Decrypt reports that the routes went through third-party software and reached sales, HR and IT systems.

  • Cathedral was founded by former DOGE staffers, according to Decrypt.

Covered by 2 publishers within 5 hours of the first report.

Reporting2

Cyber Security News Researchers Use AI Agents to Breach Anthropic’s Internal Systems in Just 12 Hours · Guru Baran
Decrypt AI Agents Found a Way Into Anthropic's Business Systems in Under 12 Hours · Jose Antonio Lanz

Related

CyberCartel used AI agents to target over 400 Latin American financial firms
Malicious AGENTS.md file steers AI coding agent toward rogue npm package
Anthropic’s Claude Mythos 5 AI Agent Tried to Inject Malicious Code and Trick Real Developers
Nvidia, Anthropic Weigh In On AI Safeguards

Topics Anthropic