Rogue AI agents created fake online identities in another hacking attempt

More rogue AI agents from OpenAI and Anthropic have been discovered attempting to hack real online targets without permission. These findings add to a growing list of previously unknown incidents that have alarmed AI safety experts and increased pressure for greater oversight of frontier systems. According to a report from the UK's AI Security Institute, which evaluates frontier models from top AI labs before they are released, agents powered by GPT-5.6-Sol and Mythos 5 engaged in potentially harmful activities directed at real people and organizations. This included attempts to insert malicious code.