Wikimedia Links OpenAI's Rogue Agents to Wikipedia Edits and May Outage

The Wikimedia Foundation reports that it has discovered activity from "rogue" OpenAI agents on Wikimedia platforms, according to a recent blog post. The activity includes edits to Wikimedia wikis, unsuccessful attempts to exploit the Etherpad note-taking tool hosted by the foundation, and heavy traffic that may have contributed to a partial outage in May.

The foundation says the wiki edits did not appear on pages visible to general readers; nearly all were testing edits made in "sandbox" areas of the wiki. It also identified a few edits to the configuration for a citation tool, which it believes were potentially malicious. The agents reportedly made millions of automated API requests.

Despite the suspicious activity, the Wikimedia Foundation states it found no evidence that its systems were used for coordination among AI agents, or that any systems or data were compromised. The disclosure follows a series of recent reports about AI agents accessing third-party websites and services without authorization.

Read More at the original source →