Spanish watchdog reports first AI agent data breach

Spain's data protection authority documented the first known data breach involving an AI agent, which used a large language model to access a system, modify personal data, and steal invoices. The report was published on September 15, 2026.