OpenAI launches Daybreak AI tool to spot and fix code flaws early

OpenAI has launched Daybreak, a cybersecurity initiative that combines frontier AI model capabilities with Codex Security to help organisations identify and patch vulnerabilities before attackers exploit them. According to OpenAI, Daybreak blends AI intelligence with Codex as an agentic harness and aims to bring secure code review, threat modelling, patch validation, dependency risk analysis, detection, and remediation guidance into the everyday development loop.

Like Anthropic’s Mythos, the goal is to tilt the defender–offender balance by detecting security issues early and addressing them before they are found by bad actors. Access to the tooling remains tightly controlled for now, with organisations encouraged to request a vulnerability scan or contact OpenAI’s sales team.

Daybreak uses Codex Security to build an editable threat model for a repository, test vulnerabilities in an isolated environment, and propose fixes, all within three model foundations: GPT-5.5, GPT-5.5 with Trusted Access for Cyber, and GPT-5.5-Cyber. Several major companies—Akamai, Cisco, Cloudflare, CrowdStrike, Fortinet, Oracle, Palo Alto Networks and Zscaler—are already integrating these capabilities under the Trusted Access for Cyber initiative.