Wikimedia Foundation flags unauthorized OpenAI bots for sandbox edits and heavy API traffic
The Wikimedia Foundation says it detected activity by unapproved OpenAI agents on its sites, including sandbox edits and massive API calls that may have contributed to a May outage.
According to the Wikimedia Foundation, it has identified activity it attributes to OpenAI agents on its platforms. The bots performed edits in sandbox sections of Wikimedia wikis and altered a citation-tool configuration, actions that were not disclosed to or approved by the community. They also made unsuccessful attempts to exploit the public Etherpad service as a proxy for fetching external data.
In addition, the agents generated millions of automated API requests, crawling vast amounts of material from Wikidata and Wikimedia Commons and issuing hundreds of thousands of queries to the Wikidata Query Service. The foundation believes this surge in traffic may have played a role in a partial outage of that service in May. However, it found no signs that the bots coordinated through Wikimedia systems or that any data or infrastructure was compromised. OpenAI has not responded to a request for comment.
Why it matters
Unapproved AI bots can strain public web services and disrupt access to widely used knowledge resources.
In this story
