tech
April 15, 2026
OpenAI releases GPT-5.4-Cyber for vetted security teams, scaling Trusted Access programme
In short: OpenAI is releasing GPT-5.4-Cyber, a model fine-tuned for defensive cybersecurity with lowered refusal boundaries and binary reverse engineering capabilities, and scaling its Trusted Access for Cyber programme to thousands of verified defenders. The move comes a week after Anthropic restricted its more powerful Mythos model to just 11 organisations, setting up a philosophical split: OpenAI bets on broad verified access while Anthropic opts for tightly gated deployment.
TL;DR
- OpenAI released GPT-5.4-Cyber, a variant of GPT-5.4 fine-tuned for defensive cybersecurity.
- The model features a lower refusal boundary for queries related to vulnerability research, exploit analysis, and malware behavior.
- GPT-5.4-Cyber includes binary reverse engineering capabilities for analyzing compiled software.
- OpenAI is expanding its Trusted Access for Cyber (TAC) program to include thousands of verified defenders and hundreds of teams.
- TAC utilizes an identity-and-trust framework with verification tiers to grant access to more capable models.
- This release is seen as a direct response to Anthropic's Project Glasswing, which restricts access to its powerful Claude Mythos Preview model.
- OpenAI's strategy emphasizes democratized access through verification, contrasting with Anthropic's approach of severe access restrictions.
- The dual-use nature of cybersecurity AI, where capabilities benefit both defenders and attackers, is a key challenge addressed by OpenAI's verification and monitoring approach.
- Top-tier users of GPT-5.4-Cyber may be required to waive Zero-Data Retention, allowing OpenAI visibility into their usage.
Continue reading the original article