OpenAI announced the release of the GPT-5.4-Cyber model, a fine-tuned version of GPT-5.4 for cybersecurity scenarios. Under specific conditions, it reduces the restrictions on rejecting security-related requests to support advanced operations such as binary reverse engineering. The model is only available to vetted security vendors, enterprise security teams, and researchers through the Trusted Access for Cybersecurity (TAC) program. OpenAI stated that a tiered mechanism will limit the model's usage and impose additional constraints on access in low-visibility scenarios.