AI trust increasingly determines whether enterprise AI scales. As organizations move beyond pilots and put models and agents into operational systems, the question is no longer whether AI performs well in isolation. Enterprises also have to determine whether the infrastructure, data and applications underneath it can withstand cyberattacks, data integrity failures and operational disruption.

That challenge is moving cyber resilience deeper into the enterprise infrastructure stack. Security controls, data protection and recovery can no longer be treated as separate disciplines when AI systems are drawing from sensitive corporate data, interacting with applications and taking actions with limited human intervention.

Those issues will be at the center of the AI Cybersecurity and Resilience Summit, with a Dell Technologies-led flagship program Oct. 27 followed by additional industry perspectives, including InfoScale and Infinidat Ltd., on Oct. 28. The broadcasts will reach North America, EMEA and APJ audiences Oct. 28–29. TheCUBE Research analysts Dave Vellante, Krista Case, Christophe Bertrand and Scott Hebner will examine how enterprises are rethinking architecture, security and operating models as AI increases both the value of corporate data and the speed at which attackers can target it. (* Disclosure below.)

“AI projects don’t fail first on capability; they fail first on trust,” Hebner said. “Until enterprises can verify and defend outcomes, autonomy stays trapped in low-stakes use cases. This is what this summit is all about.”

Bertrand positions cyber resiliency not as a supporting layer, but as the structural backbone of enterprise AI — where data governance, protection and infrastructure converge.

“There is no trusted AI in the enterprise without a cyber-resilient IT and data foundation,” he said. “And without trusted, compliant and governed data, AI credibility collapses before it ever scales.”

Dell puts cyber resilience into the AI infrastructure stack

Dell’s role in the summit reflects how closely AI infrastructure and cyber resilience have become intertwined.

AI factories introduce new attack surfaces across model inference, training data, identity, applications and the underlying infrastructure. Rather than treating those systems as another workload protected by existing controls, Dell has been pushing security deeper into the infrastructure itself.

“AI changes the whole game,” Steve Kenniston, senior cybersecurity evangelist for portfolio marketing at Dell, recently told theCUBE. “There’s the model inferencing. There’s the model training data. There are systems where people can do things like prompt injection. There’s identity management that needs to be thought about.”

That expanded attack surface is driving a security-by-design approach across the AI stack, including supply chain protections, infrastructure controls, identity and data protection.

“At Dell, we integrate security into everything that we do, right from the supply chain, through the chips, right to the device that gets delivered to you,” Kenniston added. “Where [it] breaks down in a system like this is where security falls between the cracks.”

Dell has also been broadening the role of its PowerProtect portfolio from conventional backup toward cyber resilience management. New capabilities announced this year include AI-assisted recovery guidance, enhanced anomaly detection across PowerStore snapshots and managed detection and response for PowerScale, which stores much of the unstructured data feeding enterprise AI systems.

At Dell Technologies World, Rob Emsley, director of cyber resilience product marketing at Dell, described the data itself as the constant across both traditional computing and the agentic era.

“At the end of the day, it’s the data layer that’s still the crown jewel,” Emsley told theCUBE. “Whether or not the data has been interacted [with] by a human or a digital agent, the data is still the data.”

That makes recovery an increasingly important part of the AI security equation. Dell’s strategy centers on three principles: secure, detect and recover. The company is also emphasizing isolated cyber recovery environments that preserve clean copies of essential data and applications outside potentially compromised production systems.

“We have over 2,700 customers that have deployed on-premises cyber recovery vaults,” Emsley said. “The name of the game is to put in that vault your minimal viable company. That’s the part of the infrastructure and the data and the applications that your company needs to get back online and survive from an outage.”

AI-driven attacks change the security architecture

Agentic AI adds another dimension to the threat model because autonomous systems can interact with applications, identities and data at machine speed. That shift is already changing how theCUBE Research views security architecture.

“We used to draw the stack, and security was a layer,” Vellante said recently on theCUBE Pod. “It’s not a layer anymore. It has to be something that’s orthogonal.”

The implication is that security must cut across infrastructure rather than sit alongside it. AI systems increase the number of identities, interactions and potential failure points, while attackers can increasingly use the same automation and reasoning technologies available to defenders.

Hebner further believes that scaling AI requires rethinking architecture itself, not just tooling: “The next frontier isn’t smarter agents. It’s agents whose decisions are audit-ready by design. Trust is the architecture,” he added.

Case made a related point following theCUBE’s recent coverage of Oracle’s AI cybersecurity event. Even highly mature security programs must operate under the assumption that prevention will eventually fail.

“What agentic has done is it really has solidified the fact that even very mature security programs, at some point, will likely fail,” Case said. “When it comes to cyber resilience and business resilience, the impact that this has had is that it’s really become a part of the security conversation.”

For theCUBE’s Bertrand, cyber resiliency is not a supporting function, but a gating condition: “There’s simply no trusted AI in the enterprise without a prerequisite cyber-resilient IT and data infrastructure,” he said. “Resiliency at the core is what enables autonomy at scale.”

Data integrity remains the foundation beneath enterprise AI systems. Congruity360 InfoGov Inc., for example, has emphasized the importance of eliminating redundant and outdated data before it feeds AI systems.

“We’re seeing the early adopters leverage our technology in the reduction of infrastructure costs,” said Mark Ward, chief operating officer of Congruity360. “By eliminating anywhere from 60% to 70% of the data, by eliminating rot, we’re able to reduce the amount of AI compute and AI storage required on the backend.”

That changes the measure of cybersecurity success. Preventing attacks remains essential, but enterprises increasingly must also know which applications and data are critical, determine whether those assets remain trustworthy and restore operations quickly when defenses are breached.

Operational resilience moves closer to real time

That same shift toward faster recovery and operational resilience is also evident in the strategies of InfoScale and Infinidat Ltd., a Lenovo company.

InfoScale is pushing the resilience conversation beyond restoring infrastructure and data after disruption toward maintaining a trusted operational state while an incident is unfolding. In a recent conversation with Case, InfoScale Director of Product Management Joseph D’Angelo said the traditional approach to resilience has been rooted largely in infrastructure failures, while cyberattacks and AI are expanding both the attack surface and the potential fallout.

“We know that that has changed,” D’Angelo said. “The attack surface is growing tremendously; the exposure, the fallout, all of that is increasing exponentially — so must then the response and … the way that you envision your cyber resiliency posture.”

That shift underpins InfoScale’s focus on autonomous operational resilience, which combines application awareness, continuous protection and automated response with the goal of keeping critical operations running rather than waiting for conventional recovery to begin. It also aligns closely with the summit’s broader question of how enterprises can protect the operational state of the business as AI increases the speed and complexity of cyber risk.

Infinidat is approaching the issue from the enterprise storage and data protection layer. In an interview with theCUBE’s Bertrand, Chief Marketing Officer Eric Herzog emphasized that AI systems are ultimately dependent on trusted data, making cyber protection of both primary and backup storage increasingly important.

“It’s all about the data,” Herzog said. “The core is the data, which is why you need next-generation data protection on your primary storage, which is what AI would use. You need next-generation data protection on your backup, as well.”

Infinidat’s InfiniSafe portfolio combines immutable snapshots, cyber detection and recovery capabilities designed to help organizations identify clean recovery points and avoid restoring compromised data. Herzog has also emphasized the importance of shrinking recovery times. During an earlier theCUBE Cyber Resiliency Summit interview, he pointed to a demonstration in which Infinidat recovered 2.5 petabytes of primary storage in four seconds.

“You need to have cybersecurity on primary storage and on backup storage,” Herzog said. “If you’re not including storage in your comprehensive cybersecurity strategy, that is a huge miss.”

Recovery becomes part of AI readiness

As AI moves deeper into production, cyber resilience is becoming part of the infrastructure decision itself. Enterprises need to know not only how systems are protected, but how quickly critical data, applications and operations can be restored when defenses fail.

That puts recovery architecture, trusted data and operational continuity alongside compute, storage and networking as core elements of AI readiness. The organizations best positioned to scale AI will be those that can maintain confidence in the data feeding their systems, isolate compromised environments and restore essential services without prolonged disruption.

Those priorities will frame much of the discussion at the AI Cybersecurity and Resilience Summit, where Dell, InfoScale, Infinidat and other industry leaders will examine how cyber resilience is being built into the infrastructure supporting production AI.

TheCUBE event livestream

Don’t miss theCUBE’s coverage of the AI Cybersecurity and Resilience Summit, with broadcasts Oct. 27-28 in NA and Oct. 28-29 in EMEA and APJ. Plus, you can watch theCUBE’s event coverage on-demand after the event.

How to watch theCUBE interviews

We offer you various ways to watch theCUBE’s coverage of the AI Cybersecurity and Resilience Summit, including theCUBE’s dedicated website and YouTube channel. You can also get all the coverage from this year’s events on SiliconANGLE.

TheCUBE podcasts

SiliconANGLE’s “theCUBE Pod” is available on Apple Podcasts, Spotify and YouTube, which you can enjoy while on the go. During each podcast, SiliconANGLE’s John Furrier and Dave Vellante unpack the biggest trends in enterprise tech — from AI and cloud to regulation and workplace culture — with exclusive context and analysis.

SiliconANGLE also produces our weekly “Breaking Analysis” program, where Dave Vellante examines the top stories in enterprise tech, combining insights from theCUBE with spending data from Enterprise Technology Research, available on Apple Podcasts, Spotify and YouTube.

Guests

During the AI Cybersecurity and Resilience Summit, theCUBE’s coverage will feature discussions with enterprise information and security leaders on the architecture, policies and operating models required to enhance AI trust and reliability. Stay tuned for our complete guest list.

(* Disclosure: TheCUBE is a paid media partner for the AI Cybersecurity and Resilience Summit. Sponsors of theCUBE’s summit coverage do not have editorial control over content on theCUBE or SiliconANGLE.)

Image: SiliconANGLE