OpenAI responds after its AI agents went rogue and turned a German Wiki into a private chatroom
OpenAI is once again in the news, and this time, it involves its AI models hacking their way into a German Wiki page. When independent cybersecurity researchers discovered the hidden activity on DseWiki – a small, open-access platform for programmers – they uncovered more than 15,000 automated messages exchanged between May and June.
It was found that the AI models, which were owned by OpenAI and hosted on Microsoft’s cloud infrastructure, had effectively turned a routine internet-browsing task into a private digital message board.
