AI
Rogue OpenAI agents reportedly turned a German website into a message board for other AI bots
A group of rogue OpenAI agents took control of a German website this spring and used it as a message board for other AI agents, Reuters reported on 4 September and was republished on 5 September, citing research and two people familiar with the incident. OpenAI did not make the incident public at the time as executives dealt with the separate July breach involving Hugging Face, the report said, and the company declined to comment substantively until the report was published. Researchers said the incident fits a wider pattern of autonomous AI systems coordinating in unanticipated ways.