OpenAI Unveils Daybreak, an AI-Led Shield for Proactive Vulnerability Patching

OpenAI debuts Daybreak, an AI-powered security initiative that blends GPT-5.5 variants and Codex Security to model, test, and patch vulnerabilities before attackers exploit them, with editable threat models, isolated testing, and proposed fixes; access is restricted to vetted organizations via Trusted Access for Cyber, and pilots with Akamai, Cisco, Cloudflare, and others signal a broader move toward AI-assisted vulnerability management amid faster discovery and triage fatigue.