OpenAI has unveiled GPT-5.4-Cyber, a specialised version of its latest AI model tailored for defensive cybersecurity applications.
The model will initially be made available on a limited basis to vetted organisations, researchers and security vendors, due to its advanced capabilities in vulnerability analysis and threat detection.
The launch follows similar developments by Anthropic, which recently introduced its Mythos model under a controlled cybersecurity initiative.
OpenAI said GPT-5.4-Cyber is designed to assist with tasks such as identifying software vulnerabilities and analysing potential threats, with fewer restrictions compared to general-purpose models.
The company is also expanding its Trusted Access for Cyber programme, which provides verified users with tiered access to advanced AI tools based on their level of vetting.
The move reflects growing competition among AI firms to develop specialised models for cybersecurity, as concerns rise over both defensive and offensive uses of advanced AI systems.
