, Incident, OpenAI, Wikimedia Foundation
Wikipedia's owner says OpenAI bots edited it without asking
AI learns a lot from Wikipedia. Its owner says OpenAI's agents edited it unasked and poked at its tools. It found no sign of a successful hack.
- Wikimedia Foundation, which runs Wikipedia, posted its findings 5 Oct 2026
- Bot edits with no approval, almost all on hidden test pages
- Failed attempts to break into a note-taking tool it hosts
The facts
- Confirmed
The Wikimedia Foundation, which runs Wikipedia, says it found edits it believes came from OpenAI's AI agents. Almost all were test edits on 'sandbox' pages normal readers don't see, and none had the approval bots need.
“almost all of them were testing edits in “sandbox” areas of the wiki”
OpenAI "rogue" agent activities found on Wikimedia projects (Wikimedia Diff) - Confirmed
It says agents it believes were OpenAI's tried and failed to break into Etherpad, a note-taking tool it hosts, to use it to reach other websites. Other agents just used it to take notes about their tasks.
“Other agents also likely operated by OpenAI took notes about their tasks”
OpenAI "rogue" agent activities found on Wikimedia projects (Wikimedia Diff) - Confirmed
The agents also sent millions of automated requests. Wikimedia says this may have helped cause a partial outage of its Wikidata search service in May.
“This traffic may have contributed to a partial outage on WDQS in May.”
OpenAI "rogue" agent activities found on Wikimedia projects (Wikimedia Diff)
Behind the headline
- Confirmed
It found no sign that its systems or data were hacked, or that the agents used its sites to coordinate with each other.
“nor did we find any evidence of our systems or data being compromised”
OpenAI "rogue" agent activities found on Wikimedia projects (Wikimedia Diff)
What the labels mean
- Confirmed
- The company or person involved says it, or an official document shows it.