OpenAI Launches Daybreak Cybersecurity Initiative — GPT-5.5-Cyber, Codex Security, and 8 Major Security Vendor Partners

The Verge·high signal

OpenAI launched Daybreak, a cybersecurity platform built on three GPT-5.5 variants: standard (general use), Trusted Access for Cyber (verified defensive work), and GPT-5.5-Cyber (permissive model for red-teaming and pentesting). Codex Security builds editable threat models per repository and identifies/tests vulnerabilities in isolated sandboxes. Launch partners include Akamai, Cisco, CrowdStrike, Cloudflare, Fortinet, Oracle, Palo Alto Networks, and Zscaler — OpenAI's answer to Anthropic's Mythos, framing security as a platform.