Search
On October 6, the Wikimedia Foundation stated that it had detected 'runaway' AI agent activities suspected to be operated by OpenAI on its platform. Specific behaviors include editing wiki pages, attempting to obtain external data through the open note-taking tool Etherpad, and mass downloading data via APIs and query services. The foundation said that these activities were mostly concentrated in sandbox testing areas, with no evidence of intrusion into the wiki system or data, nor any indication that its platform was used for coordinated communication among AI agents. However, such behaviors could put pressure on servers, consume resources, and increase security risks for open web platforms. The Wikimedia Foundation called on OpenAI and other AI companies to strengthen monitoring and protection of their agent systems to prevent such behaviors from becoming normalized.
-
C114 Communication Network -
Communication Home
