OpenAI Unveils GPT-5.4-Cyber: A New Model Tailored for Qualified Security Professionals

OpenAI Group PBC has launched GPT-5.4-Cyber, a specialized version of its GPT-5.4 model designed specifically for defensive cybersecurity tasks. This model features enhanced capabilities, including binary reverse engineering, allowing security professionals to analyze compiled software for vulnerabilities without source code access. OpenAI’s Trusted Access for Cyber program has also been expanded, providing access to thousands of verified security professionals. Initial access to GPT-5.4-Cyber is limited to vetted vendors and researchers, with tiered verification levels available. OpenAI’s initiative aims to prepare for future models, enhancing cybersecurity measures while preventing misuse through automated verification. Additionally, the program supports open-source security, having scanned over 1,000 projects. This announcement follows a broader trend in AI towards identity-based access controls, shifting from blanket restrictions in cybersecurity to more nuanced, targeted approaches. OpenAI emphasizes its ongoing commitment to making advanced defensive tools widely accessible.

Source link