OpenAI has unveiled two new security products under the Daybreak brand, as reported by The Decoder (DE). The first product, Daybreak Red Team, is available in early access and uses AI agents to automatically test web applications for vulnerabilities. The agents use source code and app context to examine attack paths. The tool also helps verify reported security flaws from bug bounty programs or code scanners and test patches.
Automated defense
The second product, Daybreak Defense Factory, automates defense: AI agents search companies' code for vulnerabilities and develop patches. Developers receive finished pull requests with a fix and an accompanying test for approval.
Context
According to Nancy Wang, head of Cyber-Applied-AI, Red Team is meant to deliver penetration testing at the level of automated continuous scans, but with the depth of human testers. Whether this promise holds up in practice remains open after the announcement. If you are responsible for security testing or patch processes in your organization, it is worth keeping an eye on the early access and the approval processes for generated pull requests. Both products are based on OpenAI's current cyber model, GPT-6 Cyber.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
