The future of digital infrastructure in the age of AI

The Future of Digital Infrastructure isn't Just About Capacity. At a recent Telstra executive forum, I asked a room of infrastructure and technology leaders what was driving change in their environments. AI featured heavily in the discussion, but it wasn't the only answer. What emerged was something broader: organisations are moving more data, between more locations, than ever before.

Sam Bednall, Head of Product Solutions Development, Telstra Digital Infrastructure · 23 October 2026 · 4 minute read

The architecture hasn't changed. The scale has.

Much of today's infrastructure wasn't designed around AI. It was designed to support the movement of data between applications, systems and locations.

Organisations have been dealing with growing data movement for years. Some examples are:

  • Cloud adoption has distributed workloads across multiple environments
  • Active-active applications synchronise data between sites
  • Cyber recovery platforms continuously replicate data
  • Real-time analytics requires access to larger volumes of information with increasingly demanding performance requirements

The growing adoption of AI, and increasingly, agentic AI, is accelerating demand for high-performance, low-latency connectivity. Data is moving from between users and applications to between agents, knowledge repositories, cloud platforms and operational systems. What appears to be a simple request may trigger dozens of exchanges behind the scenes before a result is returned.

More than just storing data, the challenge now is about ensuring data can move efficiently, securely and reliably between the places where business happens.

Capacity is accelerating faster than expected

The Capacity Evolution: network capacity progresses from less than 10G to 400G, spectrum services and dark fibre, supporting increasingly high-capacity cloud, hyperscale and AI infrastructure.

One of the most visible consequences of growing data movement is the demand for capacity.

Earlier in my career, moving from 1G to 10G felt like a significant step change. Today, organisations are reaching 100G and 400G requirements far more quickly as AI workloads, cloud architectures and data-intensive applications continue to grow.

The diagram above illustrates how capacity requirements continue to evolve, along with the infrastructure supporting them. What was once achievable with traditional IP networking is increasingly moving towards high-capacity optical services, spectrum and, in some cases, dedicated fibre infrastructure.

Network technology is evolving just as quickly. Earlier this year, Telstra and Ciena successfully demonstrated 1.6Tb/s transmission across a live production network, highlighting how rapidly optical networking technology continues to advance to support future demand.

As organisations become more reliant on moving data between clouds, data centres and applications, the conversation moves from bandwidth to greater certainty around performance, resilience, security and where data travels—in short, from capacity to control.

Why is control becoming more important?

This is a shift I've noticed more frequently in our customer conversations.

Questions around route diversity, operational resilience and data sovereignty are becoming more prominent. Organisations aren't just asking whether data can get from A to B. They're asking how it gets there, whether it remains protected, and whether the infrastructure supporting it can withstand disruption.

Historically, sovereignty discussions focused on where data resides. Increasingly, organisations are paying equal attention to where their data travels and the level of visibility they have over the infrastructure carrying it.

The same applies to resilience.

When critical applications, AI services and data platforms depend on the continuous movement of data, infrastructure decisions can have a direct impact on business outcomes. The network has become a strategic part of the digital ecosystem. 

What infrastructure leaders are telling us

One of the more interesting findings from the forum was that most organisations represented weren't operating in a single environment.

The majority reported a combination of on-premises infrastructure and third-party colocation facilities, while others relied on cloud platforms, their own data centres, or a mix of all three.

That means data is increasingly moving across a complex ecosystem rather than within a single platform.

It also explains why infrastructure leaders ranked resilience, governance and cost optimisation alongside AI as key priorities. These aren't separate initiatives competing for attention. They are all influenced by how effectively organisations can move, protect and control their data.

At the forum, I summarised the trend in three simple ideas.

More data. More capacity. More control

Three network infrastructure drivers: more data from AI and cloud, more network capacity to support growing volumes, and more control through strategic infrastructure.

These shifts are already playing out across many organisations today.

AI, cloud and resilience initiatives are distributing data across more locations. Capacity requirements continue to grow as organisations move larger volumes of data between applications, clouds and data centres.

At the same time, infrastructure is becoming increasingly strategic as organisations seek greater certainty around performance, resilience, security and governance.

The challenge for technology leaders is not simply keeping up with data growth. It's ensuring the infrastructure underpinning it is ready for what's coming.

Related articles