Safety overview: GPT-6 Astra
GPT-6 Astra is OpenAI's most capable model for cybersecurity.
“GPT-6 Astra is our most capable broadly deployed model and our first to reach the Critical level of cybersecurity capability under our Preparedness Framework.”— OpenAI
OpenAI has announced GPT-6 Astra as its most advanced and broadly deployed AI model, achieving a Critical level in cybersecurity capabilities. This development underscores the growing importance of AI in enhancing cybersecurity measures, marking a significant advancement for OpenAI in competitive AI technology.