Enterprises pushing AI agents into production are finding that cyber resilience now depends as much on governing what autonomous software does with data as on keeping attackers out. Much of the risk doesn’t come from malicious intent at all; it comes from agents doing unexpected things at machine speed.

That shift is pushing storage providers to build intelligent data infrastructure that can judge how data is being used. NetApp Inc. is moving access controls away from simple yes-or-no permissions and toward policies based on behavior, according to Russell Fishman (pictured), senior director of global AI solutions and field activation of NetApp.

“We’re starting to get away from thinking about everything through a human lens,” Fishman said. “This is an agent. Agents don’t think the same way. They’re not acting the same way.”

Fishman spoke with theCUBE’s Christophe Bertrand and Rebecca Knight at NetApp INSIGHT, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed a week of NetApp announcements, including Novus, a new storage architecture for AI factories and neoclouds, and the growing importance of cyber resilience as the market matures. (* Disclosure below.)

Cyber resilience moves into the data layer

NetApp’s traditional protection tools still carry weight as threats speed up. Its autonomous ransomware protection service uses AI to detect and block attacks, and a newly announced integration feeds those signals into recovery workflows from Commvault Systems Inc., Fishman explained. NetApp has built similar links with Microsoft Corp.’s Sentinel and Cisco Systems Inc.’s Splunk, he added.

“We become an indicator of compromise,” he said. “We can actually detect things and tell Commvault, ‘hey, there’s a problem,’ and vice versa.”

AI agents raise a different problem. Most organizations Fishman talks with worry less about bad actors than about agents causing unintended consequences while they try to finish an assigned task. New features in NetApp’s AI Data Engine, which prepares and governs enterprise data for AI, include a behavior-based policy engine that flags unusual patterns of data use, he noted.

“I would expect you to look at maybe one customer’s account, not all of the customers’ accounts,” he said. “That seems a bit weird to me.”

Responding at the speed of AI means organizations can’t always keep a human in the loop, so automation has to carry more of the load, Fishman explained. For NetApp customers who want stronger cyber resilience now, AI Data Engine can be deployed against any existing NetApp system, and the company announced a free six-month offer at the event, according to Fishman.

“If I was an ONTAP admin — if I was a NetApp customer — that’s the first thing I’d be doing,” Fishman said. “I want to lock this down now. The risk of not doing so, not just from a lack of innovation, but the risk of misuse of data, the lack of control, it’s too great.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of NetApp INSIGHT:

(* Disclosure: TheCUBE is a paid media partner for the NetApp INSIGHT 2026 event. Neither NetApp, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE