OpenAI Launches GPT-5.4-Cyber for Defensive Security Teams
GPT-5.4-Cyber is now available to vetted security researchers and teams defending critical infrastructure — with more permissive policies for offensive technique analysis.
OpenAI released GPT-5.4-Cyber on April 14 through its Trusted Access for Cyber (TAC) program. The model is specifically tuned for defensive cybersecurity use cases, with broader permissions than standard GPT-5 for analyzing offensive techniques, reverse engineering malware, and simulating attacker behavior.
Access: Now extended to thousands of verified individual defenders and hundreds of enterprise security teams. Organizations must apply through OpenAI's TAC portal.
Capabilities: The model has contributed to fixing over 3,000 critical and high-severity vulnerabilities through Codex Security integrations. It can analyze code for weaknesses, generate detection rules, and explain attack vectors in detail.
Debate: Some security researchers argue the vetting process is too slow and excludes solo researchers. OpenAI has committed to expanding access through academic institution partnerships.
Want automation like this for your business?
Book a free call and we'll show you exactly what's possible for your setup.