UK Safety Regulator Finds Jailbreaks That Turn GPT-5.6 Sol Into a Hacking Tool
The UK's AI Security Institute found universal jailbreaks in OpenAI's GPT-5.6 Sol that unlock autonomous vulnerability discovery and exploit development. A viral claim that Sol now beats Claude Mythos 5 on cybersecurity doesn't match the actual benchmark scores, which remain close and mixed between the two labs.