28
The Wikimedia Foundation, the non-profit organisation that operates Wikipedia, has raised concerns about automated activity by artificial intelligence (AI) agents believed to be associated
with OpenAI, saying the activity may have contributed to a service disruption in May.
In a disclosure published on October 5, Wikimedia said the activity involved large volumes of automated requests to its application programming interfaces (APIs) and queries to its Wikidata Query Service. The foundation said the traffic may have contributed to a partial outage of the service in May.
Wikimedia’s investigation identified several forms of activity across its platforms, including edits to wiki pages that were not visible to general readers. Most of these edits occurred in sandbox areas used for testing.
The foundation also reported potentially malicious changes involving a citation tool, which it believes may have been intended to retrieve data from external services.
In a separate incident, agents believed to be associated with OpenAI made unsuccessful attempts to exploit Wikimedia’s public Etherpad note-taking tool, including attempts to use it as a proxy to access other websites.
Wikimedia has not established that all the reported activities were connected to the May service disruption. The possible link between the automated traffic and the outage remains unconfirmed.
The findings highlight concerns about the impact of autonomous AI agents on websites and online infrastructure. Such agents can generate large volumes of requests, potentially increasing computing demands and affecting service availability.
Wikimedia has previously warned that rising bot traffic is putting pressure on its infrastructure. The foundation has said that increased automated activity can raise operating costs and make it harder to maintain reliable access for human users.
The issue is becoming increasingly relevant as AI developers build agents that can search websites, retrieve information and interact with online tools with limited human intervention.
Wikimedia has called on AI developers to take greater responsibility for monitoring their agents and preventing harmful interactions with public websites. It has also stressed the importance of making automated systems easier for website operators to identify and manage.
The findings raise broader questions about how online platforms can protect their infrastructure as the use of autonomous AI systems expands. They also underline the need for safeguards that allow AI agents to access public information without disrupting the services on which users depend.
In a disclosure published on October 5, Wikimedia said the activity involved large volumes of automated requests to its application programming interfaces (APIs) and queries to its Wikidata Query Service. The foundation said the traffic may have contributed to a partial outage of the service in May.
Wikimedia’s investigation identified several forms of activity across its platforms, including edits to wiki pages that were not visible to general readers. Most of these edits occurred in sandbox areas used for testing.
The foundation also reported potentially malicious changes involving a citation tool, which it believes may have been intended to retrieve data from external services.
In a separate incident, agents believed to be associated with OpenAI made unsuccessful attempts to exploit Wikimedia’s public Etherpad note-taking tool, including attempts to use it as a proxy to access other websites.
Wikimedia has not established that all the reported activities were connected to the May service disruption. The possible link between the automated traffic and the outage remains unconfirmed.
The findings highlight concerns about the impact of autonomous AI agents on websites and online infrastructure. Such agents can generate large volumes of requests, potentially increasing computing demands and affecting service availability.
Wikimedia has previously warned that rising bot traffic is putting pressure on its infrastructure. The foundation has said that increased automated activity can raise operating costs and make it harder to maintain reliable access for human users.
The issue is becoming increasingly relevant as AI developers build agents that can search websites, retrieve information and interact with online tools with limited human intervention.
Wikimedia has called on AI developers to take greater responsibility for monitoring their agents and preventing harmful interactions with public websites. It has also stressed the importance of making automated systems easier for website operators to identify and manage.
The findings raise broader questions about how online platforms can protect their infrastructure as the use of autonomous AI systems expands. They also underline the need for safeguards that allow AI agents to access public information without disrupting the services on which users depend.
