Scaling Trusted Access for Cyber with GPT‑5.5 and GPT‑5.5‑Cyber (openai.com)

OpenAI has announced the rollout of GPT-5.5-Cyber in limited preview, aimed at enhancing cybersecurity workflows for defenders securing critical infrastructure. This follows the recent release of GPT-5.5, which has facilitated powerful cybersecurity capabilities through its Trusted Access for Cyber (TAC) framework. The initiative seeks to democratize AI-powered defenses by ensuring that advanced cyber capabilities are only accessible to vetted defenders, thus allowing for tasks like vulnerability identification, malware analysis, and secure code reviews. The significance of this release lies in its dual approach to safeguarding cybersecurity: GPT-5.5 serves as a general-purpose model for a range of defensive tasks, while GPT-5.5-Cyber offers more permissive behavior for specialized workflows like red teaming and penetration testing, necessitating stronger verification measures. With enhancements designed to balance accessibility and safety, the initiative aims to empower cyber defenders, helping them to protect the ecosystem more effectively. Overall, this structured rollout aligns with ongoing efforts to incorporate advanced AI models into real-world cyber defense operations while mitigating risks of misuse.