OpenAI Says New GPT-6 Cyber Model Hunts Flaws Without Crossing the Line

OpenAI has introduced a new cybersecurity model, GPT-6 Cyber, which is rated Critical for cybersecurity. This model outperforms its predecessor, GPT-5.6 Cyber, in terms of exploitation capabilities. Additionally, GPT-6 Cyber is more reliable at staying within scope, indicating improved control over its actions. This suggests that the new model can effectively hunt for flaws while minimizing the risk of crossing boundaries. The GPT-6 Cyber model represents an upgrade over the previous version, offering enhanced cybersecurity capabilities.