OpenAI Just Launched a New AI Cybersecurity Model as AI-Powered Attacks Surge
The cybersecurity world is entering an uncomfortable new phase. AI is no longer just helping security teams write reports, analyze malware, or spot suspicious activity. Increasingly capable AI agents are being tested on complex cybersecurity tasks, and some recent evaluations have shown just how far these systems can go. Now, OpenAI is responding with a new weapon for the other side of the fight. The company has expanded its Daybreak cybersecurity program and introduced GPT-5.6-Cyber