OpenAI has introduced two new tools to deal with security problems. The newly announced tools, named Daybreak Red Team and Daybreak Defense Factory, are designed to find security weaknesses in systems and help fix them. The AI giant announced these two tools on November 12 at OpenAI's Cyber at the Frontier event.
According to the announcement, Daybreak Red Team focuses on finding security flaws in websites and apps. It can also review reports from security tools and bug bounty programs to confirm whether the problems are real.
Daybreak Defense Factory focuses on fixing the vulnerabilities identified by the first tool. Once a problem is found, its AI agents study the code and suggest a fix. The tool can also work with reports from services like Wiz, Snyk, and Endor Labs.
It can check new code and run regular security tests. However, the suggested fixes still need to be checked and approved by human experts. OpenAI says the tools are designed to work in safe settings and only test systems that have been approved.
Routine security checks may not identify vulnerabilities that could later become serious security threats. Instead of waiting for a yearly security test, companies could check their systems whenever they make changes.
Also Read: OpenAI Cites Policy Breaches in Dismissal of Three Safety Researchers
