OpenAI has broadened its Daybreak program with a new purpose-built model, GPT‑5.6‑Cyber, aimed at giving vetted defenders frontier offensive-security capabilities before threat actors can weaponize AI at scale. Daybreak Blue grants access to general-purpose frontier models, including GPT‑5.6 Sol, with safeguards tailored for authorized defensive work such as vulnerability discovery, secure code review, malware analysis, […]
The post OpenAI Expands Daybreak With GPT-5.6-Cyber for Trusted Offensive Security Research appeared first on Cyber Security News.