AI cybersecurity requires governing what trusted software does as well as keeping attackers out. As AI agents act across enterprise systems on behalf of users, identity must establish who or what is acting, how much authority it carries and where that authority ends.
That change is pushing identity providers past traditional access control and into runtime enforcement for AI agents. Enterprises also need to understand what agents have done when something goes wrong, according to Harish Peri (pictured), senior vice president and general manager of AI security at Okta Inc.
“If something does go wrong, are you able to explain and prove the series of actions that an agent took? Are you able to find all that in your logs in a single unbroken chain?” Peri said. “This is, of course, sort of in the worst-case scenario, but you have to know what happened and you have to be able to show that chain of custody in order to then fix if something does go wrong.”
Peri spoke with theCUBE Research’s Krista Case for the AI Cybersecurity and Resilience Summit, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed AI cybersecurity and what an identity architecture built for AI agents requires. (* Disclosure below.)
Building AI cybersecurity on agent identity foundations
Okta frames agent maturity as a progression of questions in its blueprint for the secure agentic enterprise: Where are my agents, what can they do, what are they doing and how do I respond? Those answers build toward a harder question — whether an agent should be allowed to take a given action at a specific moment, according to Peri.
“The real risk with agents that we’re seeing is not that [they’re] going to get hacked per se. I think that makes for good headlines, but the reality is more of the risk comes from an agent being misaligned and without any bad intention, taking an action that was never intended by the original user,” Peri said.
Answering that question starts with giving every agent its own identity. An agent that simply impersonates a user is little more than overpriced robotic process automation, while overloading a service account is a quick fix that breaks down as deployments grow, Peri explained.
“Step one is don’t take a shortcut on the foundations. If you do, it might feel like, ‘Hey, I’m innovating.’ But what will happen is, once you get to more than 10 or 15 enterprise agents that have access to your enterprise applications, you will end up having to unwind that,” Peri said.
Authority also has to travel safely as agents work across applications. Peri described a procurement agent that touches SAP, accounts payable and customer relationship management systems for users with very different permissions. The answer is delegation with downscoping, backed by central authorization and centralized logging, he explained.
“In the context of a specific prompt, the agent will never exceed the maximum allowable scope of all the parties in the chain,” Peri said. “That means that even though in theory, this agent can take actions on behalf of a manager, if it’s being invoked by frontline [personnel], it will never exceed their scope.”
Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the AI Cybersecurity and Resilience Summit:
(* Disclosure: TheCUBE is a paid media partner for the AI Cybersecurity and Resilience Summit. Sponsors of theCUBE’s event coverage do not have editorial control over content on theCUBE or SiliconANGLE.)
Photo: SiliconANGLE
Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.
15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network
SiliconANGLE Media is a recognized leader in digital media innovation, uniting breakthrough technology, strategic insights and real-time audience engagement. As the parent company of SiliconANGLE, theCUBE Network, theCUBE Research, CUBE365, theCUBE AI and theCUBE SuperStudios — with flagship locations in Silicon Valley and the New York Stock Exchange — SiliconANGLE Media operates at the intersection of media, technology and AI.
Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.
——————————————————-
Click Here For The Original Source.
