Anthropic’s security breached in 12 hours by AI startup founded by ex-DOGE staffers

A cybersecurity startup founded by former employees of the United States Department of Government Efficiency (DOGE) reportedly breached sensitive internal systems at artificial intelligence company Anthropic in less than 12 hours.

According to The Washington Post, researchers at Cathedral exploited vulnerabilities in third-party business software used by Anthropic, the developer of Claude AI.

Using AI-powered agents, the researchers accessed internal information, including meeting transcripts, salary details, job offers and Slack communications discussing advanced AI models.

The operation was conducted under Anthropic’s cybersecurity testing programme, which allows external researchers to identify vulnerabilities.

Anthropic confirmed that Cathedral accessed “a limited set” of its sales, human resources and IT systems.

However, the company maintained that customer conversations, AI model code and model weights remained secure.

“Our monitoring flagged some of Cathedral’s activity before they reported it to us, and we shut off their access the same day,” Anthropic stated.

The incident has raised concerns about cybersecurity vulnerabilities within leading American AI companies amid growing global technological competition.