Spain Reports First Personal Data Breach Involving an AI Agent

Spain’s data protection agency has reported the country’s first personal-data breach involving an agentic artificial intelligence system. Francisco Pérez Bes, president of the Agencia Española de Protección de Datos (AEPD), disclosed the incident in a 14 September 2026 post. He said an agent using a “known language model” scanned “generic files”, enabling it to log in successfully. Once inside, the agent autonomously searched the application for vulnerabilities, then used them to modify personal data and access invoices.

The AEPD has not yet disclosed further details while it investigates the breach notification. Its initial account indicates that the agent was used as an instrument to chain together several stages of an attack, suggesting deliberate use by a threat actor rather than an AI system acting independently. Simon Phillips, CTO at CybaVerse, said the incident could indicate that the attacker had jailbroken an advanced model or bypassed its safeguards, although this has not been confirmed.

Pérez Bes said AI-assisted and AI-driven attacks should be included in data-processing risk assessments, with response times reviewed to reflect the potential speed of such attacks. The AEPD also highlighted the importance of protecting digital identities and credentials, responding at machine speed, limiting access, minimising data, fixing vulnerabilities, controlling suppliers and maintaining preparedness.