AI has quickly become the defining force reshaping telecoms.

It is influencing how networks are run, how customers are served, and how operators think about future growth. While AI is often seen as intensifying network demands, it is in fact shining a light on a more familiar challenge: legacy infrastructure that is already strained, fragmented and in some cases past end-of-life.

That distinction matters. A network built on decades-old systems cannot be made resilient simply by layering in more intelligence.

In fact, if operators rush into “AI-first” strategies without first addressing the condition of their core infrastructure, they risk exposing weaknesses rather than solving them.

Recent large-scale service disruptions have made this risk tangible, reinforcing that service reliability remains the industry’s most critical measure of performance.

Ultimately, customers judge their providers on fundamentals - whether calls connect, data remains available, and services stay consistent when they matter most.

The central question is not whether AI increases network demands. It already does. The real question is whether operators have built infrastructure capable of absorbing and sustaining those demands.

Reliability remains the customer test that matters most

For telco customers, reliability is still everything. It is measured in access, consistency, and cost. When service is disrupted, the impact is immediate: trust erodes, churn risk rises, and brand reputation suffers. Telecom operators are expected to innovate, while also ensuring the uninterrupted connectivity users rely on in the modern digital age.

A recent report from Tata Consultancy Services on the state of AI in telecoms reflects this reality. AI adoption is moving from pilot to scale, with 48 percent of telecom operators now scaling AI deployments beyond pilot stages.

But despite strong momentum, the biggest barriers are structural: 62 percent identify data quality and integration issues, and 47 percent cite lack of scalable infrastructure.

This highlights that AI is not limited by the technology itself. Rather, progress depends on how ready the underlying environment is to support and sustain it.

Modernizing the core comes before scaling the promise of AI

Rather than chasing the latest trend, operators must prioritize the modernization of core systems, stabilize their data environments, and strengthen observability across the network, providing a more effective route to resilience.

This is especially true where infrastructure has become obsolete. If critical systems cannot be patched, monitored, or integrated properly, the risk of failure rises sharply. In that situation, AI cannot provide a complete solution. It can help surface issues earlier, but addressing underlying technical debt is essential to strengthening long-term stability.

That is why the most effective AI strategies in telecoms are not “AI-first” in the abstract. They are infrastructure-first, with AI used selectively to improve predictability, accelerate testing, reduce downtime, and support operational decisions. The goal is not to automate fragility but to reduce it.

AI can strengthen resilience, but only if the data and systems are ready

The good news is that AI does have a meaningful role to play in telco resilience. Networks generate enormous volumes of operational data. AI can help process that information at speed, detect anomalies, identify patterns, and flag early warning signs before they become outages.

According to that same TCS report, this is already where telcos are directing investment. 81 percent are using AI to strengthen networks and operations, closely followed by customer experience, with 73 percent citing modernizing legacy systems as the key focus for AI use. This signals a clear shift in the industry: from experimentation towards practical use cases that improve reliability and service assurance.

The next stage of this evolution will depend on the underlying environment. AI-driven operations need clean data, interoperable systems, robust governance, and infrastructure that can support real-time decisions. This is why legacy modernization matters so much, not just as a technical task but as an enabler of better customer outcomes.

Why modernization drives stronger customer engagement

When telcos modernize first, the benefits extend far beyond IT efficiency. They create conditions for better service, more consistent performance, and greater customer trust.

A more current network architecture makes it easier to manage demand surges, recover from incidents, and introduce new digital services without destabilizing the core. Better integration between systems improves visibility across the customer journey, reducing friction in everything from provisioning to support. Cleaner data makes interactions more personal, more responsive, and more accurate. And more resilient infrastructure gives operators the confidence to innovate without compromising service quality.

This is where AI can have its most lasting effect. Not as an add-on layered over an old model, but as an intelligent enabler of a better operating foundation. Used in this way, AI supports human teams, improves service predictability, and helps telcos build stronger, more trusted relationships with customers.

That is especially important at a time when industry expectations are rising, and coverage remains a core economic issue. Mobile coverage is not just a consumer issue; it is an economic one. Reliable networks underpin business productivity, digital services, and everyday connectivity. In that context, customer engagement is won through consistency, not novelty.

The telco winners of the AI era will be the disciplined ones

The telecom operators best positioned for the AI era are not necessarily the ones making the most visible claims about autonomy or transformation. They are the ones making disciplined investments in the infrastructure that supports innovation.

That means modernizing legacy systems, improving data quality, addressing patching and security gaps, and embedding AI into the network where it can genuinely strengthen reliability. It also means recognizing that AI is not a shortcut to resilience. AI can be a powerful tool, but only when the ground beneath it is stable.

The real opportunity for telcos is not to become AI-first overnight. It is to become infrastructure-smart, operationally resilient, and customer-focused, then to use AI to amplify those strengths.