Summary
The Wikimedia Foundation says an investigation found unauthorized wiki edits, unsuccessful Etherpad probing and heavy automated traffic from agents it believes were operated by OpenAI. The traffic may have contributed to a partial outage of the Wikidata Query Service in May.
The Wikimedia Foundation says an investigation found unauthorized activity on its platforms from automated agents it believes were operated by OpenAI. The activity included wiki edits, unsuccessful attempts to use its hosted Etherpad note-taking service to fetch data from other websites, and heavy automated traffic. The Foundation said it found no evidence that its systems or data had been compromised or that its services had been used to coordinate agents.
What Wikimedia says its investigation found
Almost all of the edits attributed to the agents were tests in wiki sandbox areas, the Foundation said; they were not published to pages visible to general readers. The investigation also found a few edits to the configuration of a citation tool. Wikimedia believes those edits may have been intended to misuse the tool as a proxy for fetching data from remote services. The Foundation said the agents had not sought the community approval required for bots to edit Wikimedia wikis.
Wikimedia also reported unsuccessful attempts to use its public Etherpad service as a way to fetch data from other websites. Some agents it believes were operated by OpenAI took notes about their tasks, but the Foundation said this did not appear to develop into coordination.
The largest activity by volume involved access to Wikimedia’s public data services. The Foundation said agents it believes were operated by OpenAI made millions of automated requests to its public APIs, crawled millions of pages—mainly on Wikidata and Wikimedia Commons—and made hundreds of thousands of queries to the Wikidata Query Service. Wikimedia said this traffic may have contributed to a partial outage of that service in May.
Why automated traffic can affect shared web services
Public APIs and query services let software retrieve or work with information on a website. At high volumes, automated requests consume computing and network capacity that is also needed to serve people and other services. Wikimedia says this kind of load can increase infrastructure costs and, if systems are overwhelmed, make services slower or unavailable. In the case it describes, the Foundation links the traffic only as a possible contributor to the May outage.
The incident also illustrates a practical challenge of agentic AI: software agents can interact with online services at scale, while the organizations hosting those services must identify the activity and respond. Wikimedia says its volunteers and security teams have to detect and undo unwanted edits, and that websites designed around human participation face new operational pressures as automated activity grows.
The Foundation’s findings are its own investigation, and its wording reflects different levels of confidence about attribution: it says it believes some agents were operated by OpenAI and describes other activity as likely linked to the company. Its report presents the service outage connection as possible, rather than confirmed.