[CLSTR]
API
< Back to all clusters
[TECHNOLOGY] · United States · 3 sources

started · updated

Anthropic business systems breached by AI agents in 12 hours

Anthropic Cathedral AI agents cybersecurity breach third-party software vulnerabilities

Researchers from the cybersecurity startup Cathedral successfully accessed parts of Anthropic’s internal business systems in less than 12 hours using autonomous AI agents. The breach occurred through vulnerabilities in third-party business software rather than Anthropic’s core AI models.

According to reports, the AI agents exploited flaws in administrative, sales, and human resources applications. The accessed data included meeting transcripts, employee salary and job offer details, internal Slack discussions regarding advanced AI models, and some customer and partner contact records.

Anthropic confirmed that the researchers accessed a limited set of its sales, human resources, and IT systems. However, the company stated that its core assets remained secure, noting that the researchers did not access customer conversations with Claude, source code, or AI model weights. Anthropic reported that its monitoring systems flagged the activity, and access was terminated the same day the researchers reported the findings through a bug bounty program.

Entities

Anthropic · Cathedral · Claude · HackerOne · The Washington Post

Claims

What the coverage asserts, and how many sources carry each claim.

Sources

AI Security Startup Reaches Anthropic Systems in 12 Hours [grcoutlook.com]
AI startup claims it hacked Anthropic in 12 hours, gains access to employee details and internal messages [www.techradar.com]
AI Agents Breached Anthropic's Business Systems in 12 Hours, Report Says [www.esecurityplanet.com]