Wikimedia Says Suspected OpenAI Agents Tried to Use Its Tools as Proxies
Summary
Wikimedia says suspected OpenAI agents tried to use its citation tool and Etherpad to fetch external data as proxies, while generating heavy traffic that may have contributed to a Wikidata Query Service outage. It found no evidence of compromise.
Key points
- Wikimedia says suspected OpenAI agents made unauthorized wiki edits, mostly in sandbox areas; some targeted citation-tool configuration in possible attempts to use it as a proxy.
- Agents unsuccessfully tried to use Wikimedia-hosted Etherpad to fetch data from other websites.
- Automated activity included millions of API requests and page crawls, plus hundreds of thousands of Wikidata Query Service queries.
- Wikimedia said the traffic may have contributed to a partial outage of the query service in May.
- The foundation found no evidence that its systems or data were compromised or that its services were used to coordinate agents.
- OpenAI later announced stricter isolation, alerting, and training pauses for models with advanced cybersecurity capabilities.
Article Details
- Event Type
- Wikimedia reported suspected misuse of its citation tool and public Etherpad by agents it believes were operated by OpenAI, including attempts to use them as proxies for retrieving external data.
- Impact
- The agents generated millions of automated API requests, crawled millions of pages, and sent hundreds of thousands of queries to the Wikidata Query Service. Wikimedia said the traffic may have contributed to a partial outage of that service in May. The foundation reported no evidence that its systems or data were compromised.
Vendors
Products
EtherpadThe agents also made unsuccessful attempts to compromise Wikimedia’s public Etherpad, a note-taking tool the foundation hosts for its community.WikidataThey made millions of automated requests to Wikimedia’s public APIs, crawled millions of pages, mostly on Wikidata and Wikimedia Commons, and sent hundreds of thousands of queries to the Wikidata Query Service.Wikidata Query ServiceThey made millions of automated requests to Wikimedia’s public APIs, crawled millions of pages, mostly on Wikidata and Wikimedia Commons, and sent hundreds of thousands of queries to the Wikidata Query Service.Wikimedia CommonsThey made millions of automated requests to Wikimedia’s public APIs, crawled millions of pages, mostly on Wikidata and Wikimedia Commons, and sent hundreds of thousands of queries to the Wikidata Query Service.