Across the telecoms industry, the conversation has shifted quickly towards AI-ready networks, data centre connectivity, automation, cloud-native operations, fibre expansion and next-generation infrastructure. Operators are under pressure to build networks that can support new forms of demand, from AI workloads and data centre interconnect to low-latency services, network intelligence and more automated operations.

This is an important shift, and it is easy to understand why it is attracting attention. PwC’s latest global telecom outlook describes AI and data centres as driving a new infrastructure investment cycle, with telecom operators positioned to support AI-era connectivity, sovereignty and digital infrastructure demand. Deloitte also highlights AI data centres and AI-enabled operations as major themes for telecoms in 2026.

But while the industry looks forward, many operators still face a very practical question:

What happens to the legacy infrastructure that still has to keep running while transformation takes place?

The future network still depends on the present network

Network transformation is rarely a clean break from old to new. In the real world, operators often run a complex mix of legacy, current and emerging technologies side by side.

Older SDH, optical transmission, routing, switching, access, mobile backhaul and core network platforms may no longer be the focus of new investment, but many still carry live traffic, support critical services, connect regional infrastructure, or form part of wider migration plans.

For some organisations, these platforms are still performing an important role because they are stable, paid for, operationally understood and difficult to replace quickly. For others, they remain in place because full migration is expensive, disruptive or dependent on wider commercial, regulatory or customer timelines.

That means the race towards AI-ready infrastructure does not remove the need to support legacy networks. In many cases, it makes that support more important.

Capex pressure makes the legacy support question harder

At the same time as operators are being encouraged to modernise, many are also working within tighter capital environments. Dell’Oro Group has forecast that worldwide telecom capex will decline by 2% in 2026, with only modest growth expected through 2030.

That creates a difficult balance. Operators need to invest in the future, but they also need to control spend across the infrastructure they already operate. Replacing every ageing platform immediately is often unrealistic. Extending life indefinitely without a support strategy is also risky.

This is where legacy network support becomes a commercial as well as a technical issue.

The question is no longer simply whether a piece of equipment is old. The more important questions are:

• Can it still be supported?
• Can replacement parts still be sourced?
• Can faults still be diagnosed and escalated?
• Can repairs be carried out cost-effectively?
• Is there still internal knowledge of the platform?
• Is the current support model aligned with how long the equipment will remain in service?

These questions matter because the cost of doing nothing can rise quietly over time. A platform may remain stable for years, but when a fault occurs, the lack of available spares, technical knowledge or escalation support can quickly become a business continuity issue.

Legacy decommissioning is becoming a board-level challenge

The pressure to modernise legacy environments is now being discussed at a strategic level, not just an engineering level. EY’s 2026 telecoms risk analysis highlights that the rapid transition to AI is creating organisational challenges, including the urgent need to decommission legacy IT and networks.

That is an important point. Decommissioning legacy infrastructure is not just about switching something off. It requires planning, technical understanding, asset visibility, migration sequencing, spare parts strategy, service continuity planning and commercial control.

In some cases, the right approach may be to retire the platform. In others, it may be to maintain it for a defined period while a wider migration takes place. In others, it may be to recover value from surplus equipment, redeploy usable assets, or use repairs and refurbishment to extend service life in a controlled way.

The organisations that manage this well are likely to be those that treat legacy support as part of the transformation strategy, not as an afterthought.

The hidden support gap

One of the biggest challenges with legacy infrastructure is that the support gap often appears gradually.

At first, the equipment still works. Then the OEM support model changes. Then lead times increase. Then replacement parts become harder to source. Then internal engineers with platform experience move roles or retire. Then documentation becomes harder to locate. Then a fault occurs and the organisation discovers that its escalation route is less clear than expected.

By that point, the issue is no longer theoretical.

This hidden support gap can affect operators, carriers, utilities, rail networks, infrastructure owners, data centre connectivity providers and organisations running private or specialist telecom environments. It is especially relevant where networks are multi-vendor, hybrid, regionally distributed, or built around technologies that are no longer the centre of manufacturer investment.

Modernisation does not mean abandoning legacy infrastructure

There is a misconception that legacy support sits in opposition to transformation. In reality, the two are closely connected.

A well-managed legacy support strategy gives organisations more control over how they modernise. It allows them to keep critical systems stable while they plan migrations properly. It can reduce the pressure to make rushed replacement decisions. It can help avoid emergency procurement. It can also support sustainability goals by extending the usable life of equipment, repairing rather than replacing where appropriate, and recovering value from surplus assets.

For network teams, this can mean fewer unknowns.

For procurement teams, it can mean more options.

For senior stakeholders, it can mean greater confidence that modernisation is not creating avoidable operational exposure elsewhere in the network.

What a practical legacy support strategy should include

A strong support strategy should begin with visibility. Organisations need to understand what equipment remains in service, what role it plays, how long it is expected to remain operational, and what support arrangements currently exist around it.

From there, the focus should move to the practical details.

That includes identifying critical platforms, reviewing spare parts availability, understanding repair options, mapping escalation routes, assessing internal knowledge, checking vendor dependency, and deciding where independent support may provide a more flexible or cost-effective route.

For some organisations, the priority will be urgent spare parts access. For others, it will be L3 technical escalation. Some may need repair and refurbishment support. Others may need to recover value from surplus equipment released during network upgrades.

The important point is that legacy support should be structured around the real network environment, not treated as a generic service.

Where Carritech supports this challenge

Carritech works with organisations running legacy, hybrid and multi-vendor telecom infrastructure. Our role is to help operators and infrastructure owners maintain service continuity across equipment that may be difficult to support through standard manufacturer channels.

This includes access to spare parts, repairs and refurbishment, independent L3 Remote Technical Support, asset management, deinstallation, resale and responsible recycling.

For organisations investing in new infrastructure, this support can provide a practical bridge between where the network is today and where it needs to go next. It helps ensure that legacy platforms are not ignored simply because the strategic conversation has moved on.

The AI network upgrade race is real. But so is the legacy infrastructure problem sitting behind it.

The organisations that manage both sides of that challenge will be better placed to modernise with confidence, control costs, protect service continuity and make better use of the assets already within their networks.