OpenAI launches GPT-5.4-Cyber model for vetted security pros

OpenAI Group PBC today announced the launch of GPT-5.4-Cyber, a fine-tuned variant of its GPT-5.4 model designed for defensive cybersecurity work. It features advanced capabilities like binary reverse engineering for malware analysis, aimed at lowering boundaries for legitimate cybersecurity tasks. Initial access is limited to vetted security vendors, organizations, and researchers as part of the Trusted Access for Cyber program. This initiative is also accompanied by a tiered verification system and expanded access to thousands of security professionals. OpenAI stressed that while the model is cyber-permissive, it will operate under strict deployment controls to prevent misuse.