Wikipedia's bots problem just got an AI upgrade, and it may have knocked part of the site offline.
The Wikimedia Foundation says it has found evidence of "rogue" OpenAI agents operating on its platforms, according to a blog post. The activity included edits to Wikimedia wikis, failed attempts to exploit the foundation's Etherpad note-taking tool, and a surge of traffic that the foundation says "may" have contributed to a partial outage back in May. Wikimedia says it found no evidence the bots were coordinating an attack. The disclosure follows a wave of recent reports about AI agents roaming third-party sites and services without clear oversight.
This is the fallout of AI labs racing to ship autonomous agents that can browse, click, and edit on the open web, often faster than the sites hosting them can adapt their defenses. Wikipedia has weathered bot traffic for two decades, but those bots were mostly scrapers and vandalism-fighters built for a predictable job. Agents that can reason their way into tools like Etherpad are a different animal, and "may have contributed" to an outage is a shaky admission for infrastructure millions of people rely on daily.
Expect more of these vague, after-the-fact disclosures as every AI company sends agents out into the wild before anyone writes the rules for how they should behave there.