OpenAI, 사이버 방어를 위한 ‘Trusted Access’ 확장 전략과 GPT-5.4 Cyber 공개

OpenAI is shifting the AI security paradigm from restricting model capabilities to a trust-based identity verification framework to address the dual-use dilemma. By providing defense-specialized models like GPT-5.4 Cyber to verified professionals, they aim to accelerate defensive capabilities while mitigating malicious exploitation. This strategy marks a transition toward a licensed usage model for high-risk AI technologies.