Cathedral AI agents find access paths into Anthropic systems in under 12 hours

What happened

Cathedral, a startup founded by ex-DOGE staffers, says its AI agents (AI that carries out multi-step tasks rather than answering one question) found paths into Anthropic's sales, HR, and IT systems through third-party software. Share Cathedral, a cybersecurity startup, recently revealed that its AI agents successfully identified access paths into Anthropic ’s business systems in less than 12 hours. Anthropic is an artificial intelligence company based in San Francisco, and its products and services include Claude and Claude Code.

Although the breach was detected and restricted swiftly, the incident has raised concerns over the robustness of security protocols at Anthropic, known for developing the AI model Claude. Decrypt News Artificial Intelligence AI Agents Found a Way Into Anthropic's Business Systems in Under 12 Hours Cathedral, a startup founded by ex-DOGE staffers, says its AI agents found paths into Anthropic's sales, HR, and IT systems through third-party software.

In brief Cathedral says AI agents found paths into Anthropic's internal systems in less than 12 hours by exploiting weaknesses in third-party administrative and sales software. Anthropic confirmed access to a limited set of sales, HR and IT systems, but says customer conversations, Claude responses, source code and model weights were not reached.

Anthropic says Cathedral overstated the national security implications and did not disclose all of its testing; Cathedral says it kept data access to the absolute minimum. Researchers at a cybersecurity startup said AI agents found paths into Anthropic's business systems in less than 12 hours, according to a Washington Post report published Monday.

The exercise, part of an authorized security test, uncovered vulnerabilities through third-party software used by Anthropic. Join 34,000+ readers across crypto, finance, and policy.