OpenAI Introduces GPT-6 Cyber for Authorized Vulnerability Research
OpenAI has introduced GPT-6 Cyber, a cybersecurity-focused model intended to help identify exploitable vulnerabilities and develop corresponding fixes.
The model is trained to remain within authorized scope as it assists with cybersecurity tasks. This emphasis on bounded operation reflects the broader need for AI-assisted security tools to support defensive and authorized research without exceeding the limits of their access or purpose.