OpenAI Launches GPT-5.4-Cyber for Cybersecurity Experts

On April 15, OpenAI announced the release of GPT-5.4-Cyber, a variant of its main model fine-tuned specifically for defensive cybersecurity applications. This release is part of a broader initiative called Trusted Access for Cyber (TAC), which has expanded from a limited pilot to include thousands of individually verified defenders and hundreds of teams, aimed at providing advanced AI tools to enhance cybersecurity measures against vulnerabilities.