Ankara:Unauthorized activity by automated AI agents from US developer OpenAI has been detected across Wikipedia's digital platforms, raising concerns about the impact on community infrastructure and volunteer resources.

According to Anadolu Agency, the Wikimedia Foundation reported the discovery of these "rogue" agents, which engaged in unauthorized operations such as test edits in sandbox areas and configuration changes to probe remote data fetching. The bots also made unsuccessful attempts to access the public Etherpad note-taking tool.

There is no evidence that the platform's systems or data were compromised, nor that the agents coordinated their activities. However, the automated systems linked to OpenAI conducted aggressive web scraping, making millions of automated requests to Wikimedia's public APIs, crawling pages, and querying the Wikidata Query Service extensively.

The foundation noted that this heavy traffic might have contributed to a partial outage of the query service in May. This incident adds to existing infrastructure strain, with bandwidth usage reportedly increasing by 50% since 2024 due to bot traffic, and 65% of resource-intensive activities being generated by automated systems.

Wikimedia's policies permit bot edits if reviewed and approved by the community, but no such approvals were sought for these actions. The foundation emphasized the need for technology companies to take responsibility for securing their systems and protecting the public from potential harm caused by autonomous tools. They urged AI companies to acknowledge their responsibility to monitor and prevent these risks and to operate transparently to allow website owners to manage interactions with their services effectively.