How AI Is redefining data value and why storage must keep up

April 23, 2026
How AI Is redefining data value and why storage must keep up

Security Journal UK hears exclusive insights from Samehra Malik, EMEA Marketing Lead for Video Imaging Applications at Seagate, on how data is reshaping the future of AI-driven security and storage.

Can you introduce yourself and your role at Seagate?

My name is Samehra Malik and I’m the EMEA Marketing Lead for Video Imaging Applications at Seagate, based in the UK.

My work spans across cross-functional teams where I help drive strategic programs that support our storage technologies and our customers’ evolving data needs.

A big part of my role is ensuring that complex initiatives – from product readiness to process transformation – run smoothly, efficiently and with alignment across operations and customer-facing teams.

I’m passionate about how Seagate’s innovation culture translates into real impact for our customers; especially as data becomes the foundation of the cloud and AI economy.

How has the definition of “valuable data” evolved over the past decade?

Ten years ago, “valuable data” was often defined by its immediate operational use – what was needed for analytics, backup or compliance.

Everything else was often seen as redundant or disposable.

Today, AI-enabled systems have completely rewritten that definition.

Data is no longer a by-product of digital systems; it’s a non‑fungible, irreplaceable asset.

Its value compounds over time and it fuels innovation, competitive advantage and societal progress.

In other words, data isn’t exhaustible; its potential grows the longer you keep it and the smarter AI-enabled systems become.

What risks do organisations underestimate when it comes to long-term data storage?

Many organisations still underestimate three critical risks:

  1. The gravity of data growth: AI systems generate and transform data at a scale that traditional storage architectures can’t sustain. A powerful point to make is that data has gravity; it gets heavier, richer and harder to manage every year
  2. Retention and regulatory exposure: Governments and industries are extending retention mandates and more than 150 countries now have data protection laws requiring onshore storage. Underestimating this creates compliance and sovereignty risks
  3. The cost of deleting data: In the AI era, deleting data can mean deleting future value. Many organisations still treat storage as a cost centre, not as the foundation of their AI pipeline. This leads to premature data disposal, lost insights and reduced ROI on AI investments

How can storage infrastructure keep pace with exponential AI data demands?

In today’s video imaging landscape, AI has transformed video from a passive record of events into an active, real‑time intelligence engine.

Modern security systems don’t just store footage; they analyse, correlate, detect and respond, often in milliseconds.

This shift has dramatically increased the pressure placed on storage infrastructure.

High‑resolution cameras, multi‑stream recording, longer retention requirements and real‑time analytics push data volumes and the complexity of managing them, to unprecedented levels.

To support this new era of intelligent security, storage must deliver massive throughput, ensuring that continuous streams of HD and 4K/8K video are ingested, processed and retrieved instantly.

Any lag introduces blind spots/gaps that security operations simply cannot afford.

Video imaging doesn’t sleep. From smart cities and retail environments to transportation hubs and critical infrastructure, systems must operate with 24/7 data pipelines capable of maintaining uninterrupted capture and analysis.

AI‑enabled video analytics, such as object detection, behaviour recognition and anomaly identification, depend on immediate access to large datasets and historical video archives.

Storage must perform consistently and reliably under constant load, enabling security teams to derive insights in real-time while preserving every frame for incident investigation, compliance and evidentiary needs.

This always-on environment also demands endurance at fleet scale.

A single deployment can include hundreds or thousands of cameras, each generating continuous data streams.

The drives behind these systems must withstand constant writes, simultaneous multi-stream recording, frequent data retrieval and the heavy workload profiles that define security video.

Failures carry steep consequences, not only operational downtime, but the potential loss of critical forensic evidence.

That is why next generation storage must be engineered for extreme durability, workload optimisation and long-term reliability across entire drive fleets, not just individual units.

As video retention timelines grow, from days to months or even years, storage must also deliver highly efficient economics.

AI-driven video imaging multiplies data volume through analytics, replication, higher frame rates and richer metadata.

Organisations are under pressure to scale affordably without compromising performance, availability or compliance.

Efficient storage architectures reduce total cost of ownership, optimise power usage and ensure that every investment in cameras, analytics platforms and security personnel is fully maximised.

In the modern security ecosystem, storage is no longer the silent backend of video imaging; it is the backbone that enables intelligent protection, operational awareness and trustworthy, actionable insight.

As video gets heavier, smarter and more indispensable, only storage built for heavy‑duty performance can carry the weight.

Video imaging applications don’t pause and neither can the data infrastructure behind it. Lightweight storage won’t cut it. This is heavy-duty work.

To keep pace, security infrastructures are embracing HDD‑centric performance architectures that ingest continuous multi‑stream video, often in 4K and higher, while feeding AI analytics engines without delay; this is how GPU farms stay saturated with frames for object detection, recognition and behaviour analysis instead of idling while they wait for data.

The architectural choice matters, because what once looked like a simple write‑heavy NVR now functions as a live data pipeline stretching from camera to core to archive and back again for rapid retrieval, model retraining and case review.

As deployments scale across campuses, cities and transportation networks, operators must expand capacity with hyper‑efficient economics that turn each additional terabyte into measurable operational advantage: Longer retention to satisfy evidentiary and compliance mandates, higher frame rates and bitrates for forensic clarity and richer metadata to accelerate investigations; all while keeping total cost of ownership in check so the budget goes to adding coverage rather than constantly replacing infrastructure.

Underpinning this end‑to‑end performance are purpose‑built performance hard drives engineered for the realities of security video: Relentless 24/7 writes from hundreds of cameras per array, simultaneous playback for command centres and investigators, frequent analytics passes over the same footage and harsh, always‑on environments.

Seagate’s focus on heavy‑duty durability, extreme endurance and hyper‑efficiency translates directly into predictable scale and reliable evidence integrity, so agencies and enterprises can record every frame, analyse every scene and respond with confidence – day after day, fleet after fleet.

Intelligence runs on data. Security runs on storage that can carry the load.

What challenges had to be overcome to make technologies like Mozaic viable?

Mozaic represents one of the most advanced breakthroughs in storage in decades.

Making it viable required overcoming challenges across four key areas:

  1. Nanoscale engineering: Seagate’s drives are built using some of the world’s most advanced nano‑technologies. Achieving the precision required at atomic scale for Mozaic’s HAMR-based designs was a massive scientific challenge
  2. Manufacturing at unprecedented scale: Building these technologies for global cloud and AI data centres meant retooling production to deliver extreme density, reliability and consistency at high volume
  3. Optimising for modern workloads: Mozaic had to be purpose-built for active, in‑flight, data‑intensive AI workloads, not legacy cold data use cases
  4. Reinventing fleet-level performance: Ensuring that these drives didn’t just perform individually, but acted as performance multipliers across hyperscale fleets, required new architectures and new testing paradigms

This article was originally published in the April edition of Security Journal UK. To read your FREE digital edition, click here.

Read Next

Security Journal UK

Subscribe Now

Subscribe
Apply
£99.99 for each year
No payment items has been selected yet