Scaling Trusted Access for Cyber with GPT-5.5 and GPT-5.5-Cyber

OpenAI has announced the expansion of its Trusted Access for Cyber initiative with the integration of GPT-5.5 and the limited preview rollout of GPT-5.5-Cyber. According to the company, these models are intended to help verified defenders accelerate vulnerability research, malware analysis, binary reverse engineering, detection engineering, and patch validation while protecting critical infrastructure. Trusted Access for Cyber utilizes an identity and trust-based framework designed to place enhanced capabilities into the hands of approved defenders while restricting requests that could lead to real-world harm. Vetted defenders receive lower classifier-based refusals for authorized cybersecurity workflows, though safeguards continue to block malicious activities such as credential theft, stealth, persistence, and malware deployment. Beginning June 1, 2026, individual members accessing the most cyber-capable models will be required to enable Advanced Account Security, or their organizations can attest to having phishing-resistant authentication. GPT-5.5 with Trusted Access for Cyber serves as the primary, broadly useful model for legitimate defensive tasks, handling the vast majority of security workflows. Meanwhile, GPT-5.5-Cyber is being rolled out in a limited preview to defenders securing critical infrastructure, supporting specialized dual-use workflows such as authorized red teaming, penetration testing, and controlled validation that require more permissive behavior. OpenAI is partnering with security vendors, federal and state government entities, and commercial organizations to support this ecosystem-wide deployment.