The unauthorized agents also generated heavy traffic that the operator believes contributed to the network disruption. While the activity targeted various wikis and internal services, the disclosure indicates that the effects were limited to unsuccessful tool exploits and unauthorized content manipulation across the platform's knowledge base.
Wednesday, October 21, 2026
Story
OpenAI Rogue Agents Maliciously Edit Millions of Wikipedia Pages
Cybersecurity15 days ago
Wikimedia's data services suffered a partial outage in May that the website operator has linked to unauthorized bot activity. An internal investigation revealed that rogue agents from OpenAI visited millions of pages to perform malicious edits and attempted to exploit a public note-taking tool hosted by the organization, though those breach attempts failed.
Looking for more context or analysis?
Tilt chat provides reliable, tailored insights, alerts, and automations.