Why trusted AI will define the future of telecom resilience

Resilience in telecom used to mean keeping the network up. Today, that is no longer enough. As artificial intelligence (AI) takes on a bigger role in how networks are managed, operators must keep services seamless even as infrastructure becomes more distributed and data becomes more sensitive. That is exactly the challenge reflected in this year’s World Telecommunication and Information Society Day theme, “strengthening resilience in a connected world.” Resilience now has to be built into every layer of connectivity, from terrestrial networks and submarine cables to satellites and data centres.

That challenge is only becoming more urgent as telecom networks grow in scale and complexity. Ericsson predicts that global mobile data traffic will reach 280 exabytes per month by 2030, while in India, 5G subscriptions are expected to hit 1 billion by 2031. More traffic, more distributed infrastructure, and more AI-driven decision-making mean resilience is no longer just a back-end engineering issue.

In this environment, telecom resilience will increasingly depend on three things: sovereign control of data, intelligent orchestration across more complex networks, and the ability to operationalise AI safely at scale.

 Sovereign control of data

Telecom resilience starts with data sovereignty, because operators cannot build resilient networks if they do not know where sensitive data is flowing, who can access it, or how it is being governed. That challenge is becoming more urgent as telcos face growing pressure to expose network data for new services, ecosystem partnerships, and monetisation opportunities.

As operators expand into AI services, network application programming interfaces (APIs), and more distributed forms of inference, sovereignty is becoming as much an architectural concern as a compliance one. Operators need to understand how data moves, where it is processed, and how it is shared across increasingly complex environments.

They also need control over how data is accessed and used in ways that support innovation rather than slow it down. This is where hybrid data and AI platforms can help operators govern data access more consistently through capabilities such as data lineage and a more unified data fabric for secure data exposure.

The stakes are even higher in a more uncertain geopolitical environment, where questions around where data travels, and whether it passes through potentially hostile environments, have become part of the resilience equation.

Intelligent orchestration across increasingly complex networks

Resilience also depends on operators being able to coordinate increasingly complex, distributed network environments without compromising the user experience. A key part of future resilience will be tighter integration between terrestrial and non-terrestrial networks, with satellites adding an important fallback layer when environmental conditions or backhaul disruptions affect connectivity.

But technical integration alone is not enough. What matters is whether that complexity can be absorbed without disrupting the user experience. Users do not care which network they are connected to; they simply expect the service to work. That is where AI becomes important, helping operators manage mobility, signaling and real-time switching across network environments so connectivity remains seamless.

This matters because the operational strain is already showing. According to Cloudera’s latest Data Readiness Index, three out of five telecommunications respondents said infrastructure performance consistently hinders operational initiatives, the highest among all industries surveyed. As new satellite constellations become part of the connectivity landscape, resilience will increasingly depend on how intelligently operators can orchestrate a more distributed network environment.

That is also where the opportunity lies: enabling distributed data infrastructure that can support low-latency, private AI applications closer to the network edge. In that sense, AI’s role is not to add another layer of complexity, but to act as the coordination layer that helps networks adapt in real time.

Operationalising AI safely at scale

Building resilience in the AI era requires operators to operationalise AI safely, because networks cannot become more resilient if the systems making decisions inside them are still difficult to govern, monitor, or trust.

Most telcos are still early in their AI maturity journey, and that caution continues to shape the pace of adoption. Operators remain highly risk-averse, with many still concerned that introducing AI into the network could disrupt services they cannot afford to break. That is why the bigger challenge is not adopting AI in principle, but building the operational foundations needed to use it responsibly at scale.

That means putting the right data workflows, governance models, monitoring practices, and accountability mechanisms in place before AI is pushed deeper into live network operations. According to a KPMG report, in India, 97 per cent of telecom operators are adopting or assessing AI and nearly half have embedded it into daily operations. However, they still witness the complexity of integrating autonomous AI systems, the need for robust AI governance, and the imperative to upskill workforces to collaborate effectively with intelligent agents. The issue is not lack of ambition but whether the foundations are strong enough to support AI in production.

That gap is especially visible in agentic AI. Although momentum is growing and the AI-RAN alliance showcased a number of impressive demos at MWC 2026, the conversation remains rooted in experimentation, human oversight, and unresolved questions around guardrails and sovereignty. Telcos may want the benefits of AI-native operations and agentic automation, but resilience will ultimately depend on whether they can scale these systems with the right oversight, explainability, observability, and accountability in place.

As networks get smarter, they also get harder to govern. More automation, more distributed infrastructure, and more data moving across more environments can make telecom more adaptive, but also more fragile if the foundations are weak.

That is why resilience has to be designed into how data is governed, how networks are orchestrated, and how AI is deployed from day one. For telecom operators, the real opportunity is to build networks that remain trusted, controllable, and resilient as intelligence becomes more deeply embedded across the system.