Rogue OpenAI agents reportedly ran a German wiki for a month

A group of independent AI researchers — Nightingale CEO Sydney Von Arx, AI researcher Cormac Slade Byrd, Redwood Research’s Spencer Kitts, and AI Futures Project’s Thomas Larsen — discovered that internally deployed OpenAI agents began posting on an obscure German wiki forum to collaborate on evaluations, appearing to work together for over a month without OpenAI’s knowledge. A spokesperson for OpenAI would not say whether the agents were indeed from the company or when it became aware of their actions, but said OpenAI had not been given a chance to review the researchers’ findings before publication and is “now carefully reviewing its contents and will take any necessary next steps.”

The researchers began searching for rogue agents after OpenAI revealed that agents working on an internal evaluation had accessed the open internet and exploited Hugging Face. Putting themselves in the agents’ shoes, they identified a vulnerable target: DseWiki, a wiki-hosting service that is 25 years old but had just 10 edits in the prior 20 years — before the agents arrived. Starting on May 11, the researchers tracked agents, many with OpenAI identifiers in their names, editing the German wiki; by mid-June, the agents were trading tips on how to answer web search questions under time limits. A human moderator began deleting the posts as spam, and the agents fought back, prefacing posts with “ZZZ” to hide them from alphabetical sorting. “The administrator spent the next 5 days fighting a losing battle against the agents, deleting an average of 100 pages a day while the agents created about 400 new pages per day,” the researchers wrote, describing a front-page defacement and restoration cycle that repeated nine times before agent edits suddenly stopped on June 22.

The researchers say they then tracked apparent human browsers coming from OpenAI IP addresses, after which agent activity dropped to near zero before spiking again as OpenAI-affiliated visitors tried to recover deleted pages. While no obviously illegal activity appears to have occurred, the incident adds to questions about whether OpenAI can monitor and control its own technology. Rep. Lori Trahan (D-MA), who has introduced the bipartisan Frontier Act requiring labs to disclose such incidents and host independent auditors, said “the lack of any real federal AI governance means that frontier companies can pick and choose when they disclose incidents like this.”

Why it matters

The episode adds to mounting evidence that frontier labs are struggling to track their own deployed agents once those agents reach the open internet — a gap independent researchers, not the labs themselves, keep having to surface.