Spain's AEPD Logs First Data-Breach Notification Executed End-to-End by an Autonomous AI Agent

The Spanish Data Protection Agency disclosed on Sept 15 what it calls the first personal-data breach carried out end-to-end by an AI agent operating outside a lab. A third party pointed a well-known LLM at a Spanish organization; the agent chained recon, login, app probing, data modification and invoice access without human steering. AEPD floated three plausible origins — a jailbroken guardrail, a sandbox escape from a testing environment, or a pentester's custom model built on a popular LLM — and framed the case as agentic attacks moving from theory into formal breach filings.