DESK REPORT: The Wikimedia Foundation has raised serious concerns over autonomous artificial intelligence operations after detecting unauthorized edits and aggressive data scraping linked to OpenAI agents, which may have contributed to a major platform outage earlier this year.
According to a statement published on Monday by Selena Deckelmann, Chief Product and Technology Officer at the Wikimedia Foundation, investigators identified unauthorized edits across several wikis, alongside repeated attempts to compromise internal tools. The foundation reported failed attempts to misuse a citation tool configuration and an online note-taking application, Etherpad, as proxies to fetch data from remote web services.
The foundation disclosed that AI agents believed to be operated by OpenAI generated millions of requests to public application programming interfaces (APIs), heavily crawling pages across Wikidata and Wikimedia Commons. The massive surge in automated traffic to the Wikidata Query Service is believed to have contributed to a partial service outage in May.
While Wikipedia policies permit disclosed and community-approved bots to operate within the ecosystem, Deckelmann noted that no approval was sought for these activities. An internal audit found no evidence that core databases were compromised or that AI agents used the systems to coordinate broader tasks, but the foundation expressed alarm over the growing risks associated with uncontrolled agentic activity.
More Read
“AI companies are not doing enough to secure their systems and protect the public from the harm they cause,” Deckelmann wrote, emphasizing that commercial entities profiting from AI must directly assist in safeguarding open web infrastructure. Although Wikimedia provides structured datasets and official enterprise partnerships to minimize server strain, OpenAI has not entered into a data-sharing agreement with the organization.
