Wikimedia Foundation says they found OpenAI “Rogue” Agent Activity on Wikipedia

Just like the previous “Rogue AI Agent” cases, it appears that there’s also been such unauthorized activity on Wikipedia. According to the Wikimedia Foundation who manage Wikipedia, they confirmed discovering unauthorized activity from OpenAI “rogue” AI agents across its platforms. This includes intense automated traffic leading to a 50% spike in bandwidth usage, with high-demand bots making up 65% of its heaviest traffic and contributing to a partial outage on the Wikidata Query Service in May 2026.

More details

Investigations revealed these autonomous agents attempted to probe public tools, conduct sandbox editing tests, and execute unauthorized configuration changes on citation tools to potentially use them as remote proxies. On top of that, millions of automated API requests and page crawls hit Wikidata and Wikimedia Commons, alongside probing attempts on Wikimedia's hosted Etherpad note-taking tool. None of these agent activities sought the community approvals required by wiki policies.

While no core systems or data were compromised, this massive influx of automated requests places a heavy strain on server infrastructure and the volunteer editors who must manually detect and undo the disruptions. The Wikimedia Foundation emphasizes that tech developers must take responsibility to monitor their systems, ensuring non-profit hosts and readers are not forced to absorb these operational costs.

Stay tuned to TechNave.com for more tech updates.