OpenAI launches AI security tools, warns businesses to use them ASAP or risk getting hacked

OpenAI has launched two AI-powered security tools to help businesses identify software vulnerabilities and fix them before hackers can exploit them.

In Short

• OpenAI launches Daybreak Red Team to identify and test software vulnerabilities
• It also launched Daybreak Defense Factory, which uses AI to investigate security flaws
• OpenAI says businesses now need advanced AI-powered defences to tackle AI-led cyber threats

OpenAI has launched two new artificial intelligence-powered security tools, Daybreak Red Team and Daybreak Defense Factory, as part of its Daybreak security platform. According to the company, the tools are designed to help businesses find and fix software vulnerabilities before hackers can exploit them.

The launch comes as concerns grow over AI-powered cyberattacks, with hackers potentially using more powerful AI tools to find software weaknesses and automate attacks. OpenAI warns that businesses need to rethink their security measures to keep up with these threats. With these new tools, OpenAI aims to help businesses strengthen their cyber defences by using AI to identify security gaps and fix them.

The tools use AI agents to test applications for security weaknesses, investigate potential vulnerabilities and help developers prepare fixes before hackers can exploit them.

Speaking at a keynote address on Thursday, OpenAI president Greg Brockman said companies would be secure only to the extent of the advanced AI models they use to protect themselves and how effectively they deploy them.

"We're now at a point where companies will be secure only to the extent of the frontier model that they use to defend themselves and how well they utilize that," he said, via The Register.

He further acknowledged that traditional security practices remain important but said organisations need to rethink their approach in light of what advanced AI models can do.

How will OpenAI's new AI security tools work?

Daybreak Red Team: The first tool, Daybreak Red Team, is designed to test web applications for security vulnerabilities. It uses a coordinated team of AI agents that examine an application's source code and context to investigate possible ways an attacker could break into it.

In simple terms, it works like an automated penetration-testing team. Instead of waiting for hackers to discover a weakness, businesses can use the tool to simulate potential attacks and identify security gaps themselves.

It can also examine findings from existing code scanners, bug bounty programmes and vulnerability reports to check whether the reported weaknesses are genuine. Once a company fixes an issue, Red Team can test the application again to see whether the patch actually prevents the attack.

OpenAI says businesses can control which applications and endpoints the agents are allowed to access. The tool also uses isolated environments and additional review mechanisms to keep testing within the approved scope. Daybreak Red Team is currently available through an early-access programme.

Daybreak Defense Factory: The second tool, Daybreak Defense Factory, focuses on fixing the vulnerabilities that security teams discover. It uses AI agents to examine a company's code, trace the source of a security weakness and prepare a patch for developers to review.

For instance, if a security scan identifies a weakness in a company's application, Defense Factory can investigate the issue and suggest a code change to fix it. Developers can then examine the evidence, test the proposed patch and approve it before it is added to the software.

According to OpenAI, the tool can also work with findings from existing security products and Daybreak Red Team. Companies can configure it to scan new code changes or run scheduled checks as their applications are updated.

- Ends