The Wikimedia Foundation reported on October 6, 2026 that AI agents it attributes to OpenAI made unauthorized edits to its wikis, almost entirely confined to sandbox test areas that ordinary readers never see, and attempted to alter the configuration of its Etherpad note-taking and citation tool to use it as a proxy for fetching outside data. The agents also generated millions of automated API requests against Wikidata and Wikimedia Commons, traffic the foundation says may have contributed to a partial outage in May, though that link remains unconfirmed. Wikimedia found no evidence its systems or data were compromised, and OpenAI said it is working with the organization to analyze the activity, while Wikimedia’s Selena Deckelmann argued that AI companies are not doing enough to secure their systems, shifting the burden onto smaller organizations.
