Las Vegas, October 27, 2026 / PRNewswire / -- At the Oracle AI World conference, Oracle today announced the launch of Oracle AI Security Mesh, which is supported by OpenAI. This new product complements Oracle Cloud Infrastructure ( OCI ) Identity Assurance; the latter uses biometric technology to verify real-person identities; it also works in conjunction with OCI Zero Trust Packet Routing, which protects critical systems and data by enforcing network access policies.

Phil Newman, Senior Vice President of Security and Identity at Oracle, stated: "As AI agents take on more responsibilities, enterprises need to have the security capabilities to keep up with the way these systems operate. With the help of Oracle AI Security Mesh and Oracle's identity and network security capabilities, we are helping organizations to identify and resolve security risks more quickly, while maintaining control over who can be authorized to perform sensitive operations and which systems AI workloads can access."

Leveraging Oracle AI Security Mesh supported by OpenAI to enhance corporate security

Oracle states that regardless of whether a company's systems are running on-premises, on OCI, or in other clouds, companies can leverage Oracle AI Security Mesh to enhance the protection of their applications and infrastructure. Oracle AI Security Mesh combines Oracle's cutting-edge network security models with the testing tools of OpenAI, as well as the customer-provided system context information. It can help customers identify and fix vulnerabilities by inferring application behavior, investigating potential weaknesses, or assisting in evaluating proposed repair solutions.

Based on the contextual information of each customer's system and business priorities, these models investigate how potential weaknesses could be exploited and help security teams prioritize the most critical risks. Oracle AI Security Mesh provides supportive evidence, including reproduction steps and test results, and assists teams in evaluating proposed fix solutions. Within the scope of tests approved by the customer, it tests the proposed fixes for identified vulnerabilities and checks their impact on the expected application behavior. Before applying the fix changes in Oracle AI Security Mesh, the customer must first approve the remedial measures. Each agent operates independently with its own task scope and immediate permissions, while independent monitors can detect and prevent activities that exceed the approved boundaries.

OpenAI, the person in charge of cybersecurity, stated: "Frontrunners are creating new opportunities to help organizations identify and respond to increasingly complex security threats. By combining the advanced models of OpenAI with Oracle's enterprise security capabilities, we can assist security teams in investigating complex issues, prioritizing key risks, and accelerating the speed of remediation while maintaining the necessary control measures."

Protect critical access when taking action at AI.

Oracle indicates that as companies entrust more tasks to AI, OCI Identity Assurance has introduced biometric authentication to help customers verify the authenticity of users and ensure that manual approvals come from verified personnel. This helps to protect sensitive systems and data by enhancing trust in personnel responsible for approving sensitive operations.

Before performing sensitive operations, such as making payments or changing access permissions, OCI Identity Assurance can use biometric technology to verify the identity and presence of the approver. In the AI proxy workflow, this helps to confirm that the approval comes from a verified individual. OCI Identity Assurance is built into OCI Identity, and, and Access Management, and can be applied to all Oracle Cloud services, including OCI and Oracle applications.

These biometric capabilities also help to protect the employee lifecycle, from recruitment to continuous access to corporate systems. Oracle states that OCI Identity Assurance combines identity document checks with biometric verification to help identify impostors, including those involved in North Korea's IT employee program. Subsequent biometric checks confirm that the person accessing corporate resources is the same as the registered user, helping to detect identity theft and account handover activities.

Protecting the next generation of AI workloads with built-in security in infrastructure

Oracle states that as AI workloads connect to more enterprise systems and process sensitive data, it has become more critical than ever to implement boundary controls over the content they can access. In 2024, Oracle launched OCI Zero Trust Packet Routing, taking the lead in adopting a new approach to network security and becoming the first cloud service provider to implement this emerging open standard on its platform. This capability is directly built into the network architecture of OCI, and by enforcing access policies beneath the applications and proxies themselves, it lays the foundation for protecting the next generation of AI workloads.

OCI Zero Trust Packet Routing enables organizations to express their security intentions through human-readable policies, defining which resources can communicate and which paths can access sensitive data. Traffic to protected resources must obtain permission from both the OCI Zero Trust Packet Routing policy and existing network controls. These protections remain effective even when network configurations change, helping to prevent sensitive resources from being exposed due to configuration errors. For AI workloads, this provides an independent infrastructure-level protection that helps to limit compromised workloads, prevent unauthorized lateral movement, and restrict data leakage pathways.

Oracle states that its implementation is based on the open OCI Zero Trust Packet Routing specifications established through industry initiatives, and it provides an open-source reference implementation to encourage broader adoption and interoperability. As enterprises expand the use of AI, OCI Zero Trust Packet Routing helps them to extend access to corporate data while maintaining clear and enforceable boundaries around protected resources.

About Oracle

Oracle states that the company provides integrated application suites as well as secure, autonomous infrastructure in Oracle Cloud. For more information, please visit oracle.com. Oracle is listed on the New York Stock Exchange, with the stock code ORCL.

About Oracle AI World

Oracle states that Oracle AI World is a platform for customers and partners to learn about the latest products and technological innovations, observe how AI is being implemented in various industries, and to communicate with experts and peers. Attendees will receive practical advice and insights that can have an immediate impact within their organizations, and will also learn how Oracle helps to unleash the full potential of AI.

Future Product Disclaimer

The above content is intended to provide an overview of Oracle's overall product direction and is for informational purposes only. It should not be considered as part of any contract. It does not constitute a commitment to deliver any materials, code, or features, nor should it be used as a basis for purchasing decisions. The development, release, timing, and pricing of any features or characteristics of Oracle products mentioned in this text may change at any time, and these decisions are solely at the discretion of Oracle Corporation.

Trademarks

Oracle, Java, MySQL, and NetSuite are registered trademarks of Oracle Corporation. NetSuite is the first cloud company, marking the beginning of a new era in cloud computing.