Oracle puts OpenAI models to work hunting for security holes

What's new

On October 27, 2026, Oracle unveiled AI Security Mesh at the Oracle AI World conference in Las Vegas. It is a new security product built on OpenAI's advanced cyber models and Oracle's own testing tools. The goal is to help organizations find and fix vulnerabilities, meaning weak points in IT systems that hackers can exploit.

The system analyzes how an application behaves and looks for possible weaknesses. Oracle presents it as a complement to two existing products: OCI Identity Assurance and OCI Zero Trust Packet Routing.

What's smart

Security teams get more than just a warning. The system provides reproduction steps, meaning a recipe for how to trigger the flaw, and test results as documentation. That makes it easier to tell whether a finding is real. The system also helps evaluate proposed fixes, but only within a test environment that the customer has approved.

The customer must also approve changes before AI Security Mesh carries them out. In other words, there is a human in the loop.

The other two products cover different gaps. OCI Identity Assurance uses biometrics, such as fingerprints or facial recognition, to verify the people who approve sensitive actions like releasing payments or changing access permissions. OCI Zero Trust Packet Routing enforces rules about who is allowed to talk to whom on the network, based on policies written in a human-readable language.

Phil Newman, senior vice president for security and identity at Oracle, says: "As AI agents take on more responsibility, companies need security that can keep up." OpenAI's head of cyber, Johannes Landgraf, says that advanced AI creates new opportunities to identify and respond to increasingly sophisticated security threats.

Cheaper or better?

Pricing has not yet been announced, and no date has been given for when the product will be available. It is also unclear whether it will be available in Denmark or the rest of the EU.

It is also not stated how the solution compares with competitors' security products. There are no independent tests or measurements, so all claims about what the system can do come from Oracle and OpenAI themselves.

What it's good for

AI Security Mesh is aimed at businesses and organizations with their own IT systems and security teams, not at ordinary consumers. It is especially relevant for those who already use Oracle's cloud platform OCI and want help detecting and fixing flaws faster.

Oracle points out that the need is growing as AI agents, meaning programs that carry out tasks on their own, take on more responsibility in companies. Until pricing and a launch date are available, though, it is too early to say whether the solution is a realistic choice for Danish companies.