Oracle launches AI security mesh powered by OpenAI models

LAS VEGAS - Oracle has unveiled Oracle AI Security Mesh, an AI-powered security system, in a recent press release. Developed in collaboration with OpenAI, the platform aims to assist businesses in detecting and rectifying security vulnerabilities within their systems. The novel service amalgamates OpenAI's cybersecurity models with Oracle's testing tools, analyzing application behavior to uncover potential security flaws.

Security teams are furnished with evidence such as reproduction steps and test results, and the system aids in evaluating proposed solutions within the customer-defined testing parameters, which necessitates approval before execution. Phil Newman, Oracle's Senior Vice President of Security and Identity, emphasized the increasing reliance of enterprises on AI agents, necessitating security measures capable of keeping pace with AI systems' operations.

He further stated that the Oracle AI Security Mesh, when coupled with Oracle's identity and network security capabilities, enables organizations to swiftly identify and eradicate security risks whilst preserving control over sensitive actions and AI workloads. The unveiling also incorporated Oracle Cloud Infrastructure Identity Assurance, leveraging biometric verification to authenticate user identities before enabling sensitive actions like payment releases or access permissions alterations.

This system is incorporated into Oracle Cloud Identity and Access Management and functions across Oracle Cloud services. Additionally, Oracle highlighted its OCI Zero Trust Packet Routing, introduced in 2024, which enforces network access policies at the infrastructure level, allowing organizations to outline which resources can communicate through human-readable policies, ensuring ongoing protection even as network configurations evolve.

OpenAI's Head of Cyber, Johannes Landgraf, commented on the launch, highlighting how frontier AI is creating new avenues to assist organizations in identifying and reacting to sophisticated security threats. He added that by merging OpenAI's superior models with Oracle's enterprise security capabilities, security teams can efficiently investigate intricate issues, prioritize critical risks, and hasten remediation while upholding the necessary controls. These announcements were made at Oracle AI World in Las Vegas.