Spain's Data Protection Regulator Logs Its First Breach Notification Blaming an Autonomous AI Agent for Altering Personal Data

The event

The AEPD's September 14, 2026 blog post, "Primera notificacion de una brecha de datos personales causada por un ataque ejecutado mediante un agente de IA," describes a notification in which - translated from the agency's original Spanish - "the attacking agent initiated a search for vulnerabilities in generic files, and performed a correct login," then, once inside the system, "began to search, autonomously, for vulnerabilities in the application," which allowed it to modify personal data and reach invoice records. The agency says the notifying organization identified "a well-known language model" as the tool involved, without naming it. AEPD states, in its own post, that "the information available comes from the notification presented by the affected organization and must be subject to the corresponding analysis" - meaning the agency has not itself investigated or verified any part of the account - and separately cautions that naming which AI model was reportedly used "does not imply that the model or its provider's infrastructure were compromised," nor that the tool "was designed to develop malicious activities." As quoted by SecurityWeek two days later, on September 16, AEPD framed the notification's significance as procedural rather than technical: "What is relevant from a data protection perspective is that a third party would have used an AI agent as an instrument to successfully chain together different phases of the attack," describing an agent generically as something that "can receive a goal, plan intermediate tasks, use tools, execute code, consult sources, interpret results, and modify its actions autonomously, based on what it finds." No affected organization, sector, victim count or specific AI vendor has been publicly named; by the regulator's own description, this is one unverified account of an incident that has not yet been examined. Outside commentators urged caution about over-reading it: CyberVerse chief technology officer Simon Phillips said the reporting risked "scaremongering the public with stories around AI once again running rogue," adding "we don't have enough information to understand what happened or how the model carried out this breach."