OpenAI launches GPT-5.4-Cyber for defensive cybersecurity

Key facts

• OpenAI announced a new AI model called GPT-5.4-Cyber, a cyber‑permissive variant of GPT-5.4 designed for defensive cybersecurity.
• Access to GPT-5.4-Cyber is limited to the highest‑tier users who authenticate themselves as cybersecurity defenders.
• Trusted Access for Cyber (TAC) is required to use GPT-5.4-Cyber, and it can be obtained by individual identity verification at chatgpt.com/cyber or enterprise requests through an OpenAI representative.
• GPT-5.4-Cyber is fine‑tuned with fewer capability restrictions and includes binary reverse‑engineering capabilities to analyse compiled software for malware, vulnerabilities and security robustness without source code.
• The model is being deployed iteratively to vetted security vendors, organisations and researchers as part of the expanded TAC initiative launched earlier this year.