Wikimedia Foundation, which operates the popular online encyclopedia Wikipedia, says it found activity from OpenAI’s AI agents on its platforms. The activity included unauthorised wiki edits, attempts to use a public note-taking tool to access other websites, and millions of automated requests to Wikimedia services. The company said it found no evidence that its systems or data were compromised. However, it warned that such AI agents can put heavy pressure on websites and create security risks. It also said that some of the activity may have contributed to a partial outage of its Wikidata Query Service in May.
According to Wikimedia, the AI agents made edits to its wikis without seeking the approval normally required for bots. Most of these edits were made in sandbox areas that are not visible to normal readers.
Also read: Sam Altman says AI’s benefits outweigh the bad, explains how OpenAI differs from Anthropic
It also found a few changes to the configuration of a citation tool. It believes these may have been attempts to misuse the tool to access data from external websites.
Wikimedia said the agents also made attempts to exploit its public Etherpad service, which is used as a note-taking tool. Some agents also appeared to use the service to record notes about their tasks, but Wikimedia found no evidence that its platforms were used by the agents to coordinate with each other.
Wikimedia also said OpenAI agents made millions of automated requests to its public APIs and crawled millions of pages, mainly from Wikidata and Wikimedia Commons. They also made hundreds of thousands of queries to the Wikidata Query Service.
Also read: OpenAI AI agent hacks another Australian govt department, accesses non-public data: Report
“Wikipedia was designed for humans- and agentic behaviour clearly poses challenges that no one has solutions for,” Wikimedia said.
The Foundation also pointed to the wider impact of AI-generated traffic. It said bot activity caused Wikimedia’s bandwidth usage to rise by 50 per cent in 2025 compared with the previous year. At the same time, 65 per cent of the most resource-heavy traffic on its projects came from bots.
“AI companies are not doing enough to secure their systems and protect the public from the harm they cause. That burden is falling onto everyone else, including smaller organisations. At a minimum, their systems should operate in a way that non-profit website owners like us can easily identify, and choose how they interact with our services,” the company stated.
Also read: OpenAI alerts over 100 organisations about rogue AI agent activity, check all details