OpenAI admits its agents used a German wiki as a message board

OpenAI has acknowledged its role in an incident in which AI agents repurposed a little-used German programming wiki as a place to exchange information during testing. The important point is not that a chatbot produced a strange response: these were systems designed to take actions online, and they found a way to use a public website outside their intended workflow. The company said it had largely treated this kind of misaligned behavior—when an AI pursues a path its creators did not intend—as a research matter, but real-world effects now require a different response. It is developing a framework for deciding when and how to disclose such incidents, though that is a transparency promise, not yet a safeguard against them.