How AI Is Changing Enterprise Storage Security

Guest post by Ivor Buckley, Field CTO at Dell Technologies Ireland and Northern Ireland

Enterprise data has long been a key focus for cyber threats, but AI is changing the scale, concentration and velocity of that data. As AI workloads draw on more information across the business, the storage environments that hold and protect that data are becoming even more important.

That shift is already visible in breach cost data. A study from Cybersecurity Ventures, predicts that ransomware costs will reach $74 billion annually. The risk is particularly relevant in Ireland, where cyber threats continue to intensify. According to Ireland’s National Cyber Security Centre (NCSC), ransomware remains one of the country’s top cyber risks, with increasing concern around attacks targeting critical infrastructure and interconnected supply chain.

As threats become more sophisticated and targeted, AI is both raising the stakes and changing the defence, making the integrity and recoverability of enterprise storage one of the most important questions in enterprise security.

AI changes enterprise storage security structurally. It increases the value of the data held in storage, accelerates how that data moves across systems and places new demands on visibility and control. For businesses, this makes secure, resilient and well-governed storage, a critical foundation for AI adoption.

This aligns closely with Ireland’s National Digital and AI Strategy which places significant emphasis on secure digital and AI infrastructure, strong cybersecurity posture and robust data governance as national priorities. The strategy explicitly highlights the need for resilient data foundations to support safe AI deployment across both public and private sectors.

The pressure on organisations to strengthen storage resilience is also increasing from a regulatory perspective. The EU’s NIS2 Directive, now being implemented across Ireland, places greater accountability on organisations to demonstrate robust cyber resilience, governance and recovery capabilities.

First, AI brings data together in new and powerful ways. AI systems aggregate information from across organisation into training datasets, inference pipelines and knowledge layers that unlock deeper insights. With the right storage foundation, businesses can securely harness this consolidated data to drive more accurate, real-time intelligence.

Second, AI enables more dynamic and connected data movement. Data now flows continuously between storage, compute and inference layers, creating opportunities for faster decision-making and innovation. When supported by intelligent storage visibility and control, this movement can be managed securely while delivering performance at scale.

Third, AI is accelerating the shift towards strong data governance at the storage layer. As AI adoption grows, organisations are prioritising visibility into how data is accessed and used by AI systems. With the right infrastructure controls in place, businesses can confidently govern their data, close visibility gaps and embed security into their storage architecture.

What Does Cyber-Resilient Enterprise Storage Look Like in an AI Environment?

Visibility becomes the foundation that requires continuous insight into how data moves across businesses storage environment and which systems access which datasets. Yet according to Dell Technologies’ Innovation Catalyst research, while 82% of IT decision-makers recognise that data is the differentiator for AI integration and must be used and protected accordingly, only one in three say they can turn that data into real-time insights. That gap starts at the storage layer, and without closing it, governance frameworks and policy controls cannot function effectively.

Immutable storage architecture ensures that critical data cannot be modified or deleted even by a compromised internal account.

This approach remains one of the most effective controls against ransomware in AI-adjacent environments. This was also highlighted the NESC recent paper, which states that “trustworthy and ethical practice” in AI systems reinforces the importance of resilience and integrity in underlying data infrastructure as essential to maintaining public and organisational trust in AI outcome

AI workloads should never have unrestricted access to the full enterprise data estate. Segmenting AI-accessible data from operationally critical and sensitive data, helps organisations contain the impact of a breach, limiting the risk from any single point of compromise.

Zero Trust takes that principle further, by assuming that any access points can be compromised and designs controls accordingly. In an AI-enabled storage environment, this means enforcing least-privilege access across all workloads, requiring continuous verification, rather than perimeter-based trust.

Finaly recovery architecture must be tested against AI workload requirements specifically. The question is not whether a business can recover, but how quickly it can restore the specific data dependencies that AI systems require to operate, and whether that recovery can be clearly tested, proven and repeated.

In Summary

As AI becomes more deeply embedded in the enterprise, storage security can no longer be treated as a back-end IT issue. It needs to be designed into the architecture from the start before risks emerge. For AI to scale safely, enterprises need storage foundations that are secure, resilient and governable. Businesses that address this now, will be better positioned to protect critical data, support compliance and build trust in their AI systems as business needs continue to evolve.