Refactored.
Switch Mode

Claude-powered AI agent’s confession after deleting a firm’s entire database: ‘I violated every principle I was given’

undefined 5, 2026

Claude-powered AI agent’s confession after deleting a firm’s entire database: ‘I violated every principle I was given’

Source: guim

A crisis at PocketOS shows the risks of AI automation after an AI coding agent from Cursor, powered by Anthropic’s Claude model, deleted the company’s production database and backups in seconds. Founder Jeremy Crane said the agent ignored explicit safety rules, disrupting car rental businesses reliant on the software. Data loss affected reservations and customers, forcing partial recovery from old backups. The incident highlights growing concerns that AI systems are being deployed faster than safety measures can keep up.

Read full article

Related