[Security Firm Founded by Former DOGE Member Claims to Have Breached Anthropic Systems in Less Than 12 Hours] According to The Washington Post, Cathedral, an AI cybersecurity startup founded by former members of the U.S. Department of Government Efficiency (DOGE), claims its AI agent breached Anthropic's internal systems in less than 12 hours, obtaining meeting minutes, payroll and hiring information, and Slack messages. Anthropic confirmed that the attacker accessed parts of the sales, human resources, and IT systems, including customer and partner business contact information and some basic user account records, but did not access Claude user conversations, source code, or model weights. This test was conducted under Anthropic's bug bounty program, and the intrusion was achieved using a vulnerability in third-party commercial software. Anthropic accuses Cathedral of failing to fully disclose its actions, violating program rules.