- OpenAI launched two new security products under its Daybreak brand: Daybreak Red Team and Daybreak Defense Factory.
- Daybreak Red Team, available in Early Access, uses AI agents to probe web applications for vulnerabilities by analyzing source code and mapping attack paths.
- Daybreak Defense Factory automates the opposite side of the work, scanning codebases and generating ready-to-deploy pull requests with fixes and tests.
- Both products run on OpenAI’s dedicated GPT-6 Cyber model, the same model behind OpenAI’s recent Firefox and Linux kernel vulnerability findings.
What Happened
OpenAI introduced two security products under its Daybreak brand, moving the company further into commercial cybersecurity tooling. The Decoder reported the launch on November 13, 2026.
Why It Matters
The launch builds on OpenAI’s existing Daybreak program, which the company extended to Ukraine for civilian cyber defense in September 2026, and on the GPT-6 Cyber model that separately found dozens of unknown Firefox flaws and reproduced hundreds of Linux kernel exploits.
Technical Details
Daybreak Red Team, available in Early Access, uses AI agents that analyze source code and application context to map out attack paths, verify vulnerabilities reported through bug bounty programs or code scanners, and test whether patches actually close the hole. Nancy Wang, OpenAI’s head of Cyber-Applied AI, said the tool aims to match the depth of human penetration testers “at the speed of automated continuous scanning.” Daybreak Defense Factory scans a company’s codebase and generates ready-to-deploy pull requests containing fixes and accompanying tests, which developers then approve manually.
Who’s Affected
Enterprise security teams gain an automated option for both offensive testing and defensive patching, putting OpenAI in direct competition with established vulnerability-scanning and code-security vendors.
What’s Next
Daybreak Red Team remains in Early Access, and OpenAI has not announced broader availability or pricing for either product.
