Telecom Argentina's View of Autonomous Networks and the Road to Level 4

At FutureNet World 2026, Eduardo Panciera from Telecom Argentina gave a useful operator view of where autonomous networks are heading and, more importantly, what stands in the way of getting there.

The presentation was titled Autonomous Networks and AI: A Perfect Match, From Automation to Level 4 Autonomy. The main message was simple but important. Operators cannot keep managing future networks with the same operating model they use today. Automation helps, but automation alone is no longer enough. As networks become more complex, operators need to move towards real autonomy.

Telecom Argentina is an interesting company to hear this from. It is not just a mobile operator. It provides connectivity, IPTV, OTT services, B2B services and fintech services. It has also been rebranding products under the Personal brand, with the aim of giving each user a more personal and digital experience. Eduardo linked this ambition to TM Forum's Zero-X vision, where the goal is to provide zero wait, zero touch and zero trouble experiences for customers. TM Forum describes Zero-X in similar terms, focusing on zero wait, zero touch and zero trouble as guiding principles for future digital operations.

The problem is that simplicity for the customer usually means more complexity inside the network.

Telecom Argentina's view is that operators need self-X networks, cloud-based programmable infrastructure, AI-driven assurance, automated decision-making and APIs that expose network capabilities. In other words, the customer-facing experience may become simpler, but the operational layer behind it has to become much more intelligent.

This is particularly true as 5G networks expand. 5G enables more personalised services and more programmable capabilities, but it also brings additional operational complexity. Operators have to deal with multi-cloud environments, heterogeneous networks, rising costs and more demanding service expectations. Traditional automation can support some of this, but the network is becoming too complex for humans to remain involved in every design, decision and recovery process.

That is the gap between automation and autonomy.

In a normal network operations cycle, the operator defines targets and KPIs, observes the network, identifies deviations, analyses what is happening, designs possible solutions, decides which one to apply and executes the action. Today, parts of that cycle are already automated. Execution is often automated. Observability and alarm handling are increasingly automated. But solution design and decision-making still often sit with human teams. Eduardo's point was that autonomy requires these cognitive steps to move into the system itself.

This is why Level 4 matters.

TM Forum classifies Autonomous Network levels from Level 0 to Level 5, ranging from manual management to fully autonomous networks. Its Autonomous Networks Mission describes Level 4 as introducing decision-making based on intent-driven, predictive analysis and closed-loop management of service-driven and customer experience-driven networks, supported by AI modelling and continuous learning.

In practical terms, Level 4 is where autonomous networks stop being a future vision and become an execution problem. TM Forum made a similar point in June 2026, saying that Level 4 is where autonomous networks move from ambition into an industry execution challenge.

Level 5 remains the longer-term aspiration. It implies full autonomy across a much wider range of domains and scenarios. Level 4 is the more immediate challenge because it requires operators to trust the system to make decisions in defined areas, under defined policies, using reliable data and closed-loop control.

AI is central to that transition, but Telecom Argentina's message was not simply that AI can be added to existing operations.

Eduardo described a layered autonomous network architecture, aligned with TM Forum thinking, with business, service and resource layers. The resource layer includes familiar network domains such as mobile core, RAN and transport. Intents flow down from higher layers, while reports and feedback flow up. Each domain can have its own closed loop, and there can also be closed loops between layers.

This is an important distinction. Autonomous networks are not just about automating individual tasks. They are about connecting business intent, service requirements and resource behaviour through closed-loop systems.

The AI components in this architecture can be split into copilots and agents.

A copilot is triggered by a human. It can help with suggestions, data analysis, troubleshooting and natural language interaction. It supports the human operator.

An agent goes further. It observes, analyses and decides without direct human intervention. This is where the shift towards autonomy starts to become real.

The agent model described in the presentation is also worth noting. Agents receive intent from humans or from other agents. They observe the network environment using logs, KPIs and alarms. They use knowledge and memory to understand context. Short-term memory provides the current situation, while long-term memory captures past experience, domain knowledge, design documents and technology information.

Agents can work in two ways. They can be reactive, responding to real-time events. They can also be proactive, anticipating problems or recommending network parameter optimisation before an issue becomes visible to the customer. In practice, operators will need both. Reactive autonomy helps with fault handling. Proactive autonomy is where networks start to become self-optimising and, eventually, self-evolving.

However, one agent is not enough.

Telecom Argentina's view is that operators will need a network of agents distributed across business, service and resource layers. These agents will need to collaborate, negotiate and communicate with each other. That brings a new challenge. If an operator has hundreds or thousands of agents working across domains, then it also needs a governance framework for those agents.

Eduardo highlighted several elements of this governance framework: registry, identity, guardrails, orchestration, observability and a shared data model. Agents must be able to discover each other. They must be authenticated. They must have rules and boundaries. Their workflows must be orchestrated. Their actions must be observable. Most importantly, they must all understand the network in the same way.

That final point may be the most important part of the presentation.

Telecom Argentina's argument is that operators cannot just place AI on top of the way networks are operated today and expect Level 4 autonomy to emerge.

Today, many telecom operations still work from the "how". Teams have runbooks and documents that describe how to configure the network, how to configure assurance and how to update inventory. Different teams perform different steps. The result is often fragmented data. The real network configuration may not match the inventory. The topology used by assurance systems may not match the live network. Different systems can hold different versions of reality.

In that environment, AI may improve some tasks, but it will not be reliable enough for true autonomy. Agents making decisions on top of poor or inconsistent data will make poor or inconsistent decisions.

This is why trusted data is central to autonomous networks.

Telecom Argentina's proposed shift is to move from operating from the "how" to operating from the "what". Instead of starting with runbooks and configuration steps, the operator starts with intent. What service should be delivered? What resources does it require? What SLA must it meet? What assurance should be associated with it? This information should be captured through a catalogue and a source of truth, then orchestrated across resources, inventory and assurance.

This is a different operating model.

It means that autonomy is not just about AI tools. It is about trusted data models, service catalogues, intent-based operations, orchestration and closed-loop assurance. The agents only become useful when they operate on a consistent representation of the network and the services running over it.

That also explains why Level 4 is so difficult. The technology is only one part of the journey.

Telecom Argentina described three avenues for reaching Level 4. The first is technology, moving from traditional automation to programmable networks and then to agentic AI. The second is process, because existing operating processes need to change if agents are to be used safely and effectively. The third, and perhaps most difficult, is culture. Operators have to move from thinking primarily in terms of networks to thinking in terms of data models and digital operations.

This is a useful message for the wider industry.

Many operators are now talking about autonomous networks, AI-native operations and agentic AI. The risk is that these terms become marketing labels attached to existing automation platforms. Telecom Argentina's presentation was more grounded. It made clear that autonomy requires a much deeper change in how the operator understands, models and governs its network.

For Operator Watch readers, this is also a reminder that the next phase of operator transformation will not only be about 5G coverage, fibre expansion, cloud migration or customer apps. It will also be about the operating model underneath all of that.

An operator that wants to deliver personalised digital services at scale cannot keep relying on fragmented inventories, disconnected assurance systems and manual decision-making. It needs trusted data, programmable infrastructure, intent-based orchestration and closed loops that can act safely within defined boundaries.

AI can help, but AI is not the starting point. The starting point is a trusted model of the network and the services running over it.

That may be the most important lesson from Telecom Argentina's FutureNet World presentation. Level 4 autonomy is not just a technology milestone. It is a test of whether operators can redesign their operations around trusted data, closed-loop control and governed AI agents.

Level 5 may still be aspirational, but Level 4 is already becoming the practical battleground.

The FutureNet World 2026 presentation by Eduardo Panciera is embedded below:

Related Posts: 

• Operator Watch Blog: Argentina's Mobile Market at a Turning Point 

Argentina's Mobile Market at a Turning Point

Argentina has long been one of Latin America's more unusual mobile markets. While many countries in the region have been dominated by one or two large operators, Argentina maintained a durable three-player structure, with Personal, Claro and Movistar competing nationally across mobile coverage, network quality, pricing and increasingly converged services.

According to Argentina's competition authority, before the recent market restructuring Claro had 41.8% of the national mobile market, Personal had 33.8% and Movistar had 24.4%. The split was not perfectly equal, but it ensured three nationwide network operators with sufficient scale to compete. That structure is now being reshaped.

In February 2025, Telecom Argentina acquired almost all of Telefónica Móviles Argentina, the company behind the Movistar brand, for US$1.245 billion. The transaction brought Personal and Movistar under the same ownership and immediately raised concerns about excessive market concentration. Without remedies, the competition authority calculated that Argentina would effectively be reduced to two independent mobile network operators, with the combined Telecom operation holding around 58% of the market and Claro the remaining 42%.

After more than a year of regulatory and political uncertainty, the most important development came in June 2026.

Argentina's Tribunal de Defensa de la Competencia made approval of the transaction conditional on a substantial restructuring. Telecom must transfer six million active mobile customers to an independent acquirer, including four million in the Buenos Aires metropolitan area and another two million elsewhere in the country. The transfer must be accompanied by sufficient spectrum to support a viable mobile service. The new competitor will also be able to access network sharing, roaming, co-location and other infrastructure arrangements while building its own network. The divestment must be completed within 18 months.

This means that the defining question for Argentina's mobile market is no longer simply whether Personal, Claro or Movistar has the best network.

It is who will emerge as the country's third major mobile operator.

The regulatory intervention is particularly interesting because it is not simply requiring Telecom to return spectrum or sell a small collection of assets. The objective is to preserve three competitive nationwide mobile operators by transferring a substantial customer base together with the spectrum and access arrangements required to support it.

For years, Argentina had three established mobile network operators. Telecom's acquisition of Telefónica threatened to turn that into a duopoly. The regulator is now attempting to create the conditions for a third strong competitor to emerge from the restructuring.

This change comes at an important point in the evolution of Argentina's mobile networks.

According to GSMA Intelligence data published by DataReportal, Argentina had 66.6 million cellular mobile connections at the end of 2025, an increase of 4.0 million, or 6.4%, compared with the end of 2024. This was equivalent to 145% of the population, reflecting the fact that many people and organisations use more than one connection. Around 98.4% of mobile connections were classified as mobile broadband connections using 3G, 4G or 5G networks.

LTE remains the workhorse of the Argentine mobile market, but the 5G story has been developing for several years.

Personal switched on Argentina's first 5G network in February 2021, initially activating ten sites in Buenos Aires and Rosario. These early deployments used Dynamic Spectrum Sharing, or DSS, allowing 4G and 5G to share existing spectrum. By the end of 2022, Personal had 143 active 5G sites and was gaining operational experience while waiting for the spectrum required for a larger-scale deployment.

The real turning point came in October 2023, when Argentina auctioned 250 MHz of spectrum in the 3.5 GHz range. Claro and Telecom each acquired 100 MHz, while Telefónica acquired 50 MHz. The auction generated more than US$875 million and gave the operators access to dedicated mid-band spectrum for more meaningful 5G deployments.

This distinction is important. Argentina was not particularly late to put 5G on air, but the first deployments were limited and largely based on DSS. The move from early 5G using existing spectrum to dedicated 3.5 GHz networks only accelerated after the 2023 auction.

Today, Personal, Claro and Movistar all offer commercial 5G services in major urban centres including Buenos Aires, Córdoba, Rosario and Mendoza.

However, this should still be viewed as the beginning rather than the end of deployment.

Although commercial 5G services are now available in Argentina's largest cities, LTE continues to account for the majority of mobile connections and traffic, with 5G adoption still in its early stages.

For most subscribers, 5G currently represents an incremental improvement in speed rather than a transformational change in service.

Claro entered Argentina as part of América Móvil's wider Latin American expansion and, before the recent restructuring, had grown into the country's largest mobile operator by market share.

The operator acquired a full 100 MHz block of 3.5 GHz spectrum in the 2023 auction and subsequently selected Nokia for a nationwide 5G deployment. The initial phase targeted Argentina's largest cities, using Massive MIMO and other elements from Nokia's AirScale portfolio. The deployment was positioned not only around faster consumer mobile broadband but also around future enterprise opportunities in sectors including manufacturing and oil and gas.

Claro's position has changed significantly from its earlier image as primarily the aggressive pricing challenger. It now combines the largest mobile market share with a significant spectrum holding and a major 5G investment programme.

In the old three-player structure, Claro could compete against Personal and Movistar separately. Following Telecom's acquisition, it briefly faced the prospect of competing against a much larger combined rival. The June 2026 remedies are therefore highly significant for Claro as well as for the future third operator.

Movistar is at the centre of this restructuring. Telefónica had been a major part of Argentina's telecommunications sector for decades, investing heavily in 3G, LTE, fibre and other services. It also acquired 50 MHz of 3.5 GHz spectrum during the 2023 auction.

However, Telefónica had been reducing its exposure to several Latin American markets. On 24 February 2025, Telecom Argentina acquired 99.999625% of Telefónica Móviles Argentina for US$1.245 billion. By the end of 2025, the acquired business still had approximately 19.1 million mobile subscribers, including M2M connections.

The future of those customers is now one of the biggest questions in the Argentine telecoms market.

Six million active mobile customers must be transferred to an independent acquirer. They will not simply represent a collection of disconnected assets. The regulatory package includes customer contracts, numbering, sufficient spectrum and transitional access to infrastructure. The objective is to give the acquiring operator enough scale to compete rather than creating a weak third player that exists only on paper.

The remedies extend beyond mobile. Telecom must also divest 211,400 fixed broadband customers across 28 local markets where the combined operation would create excessive concentration. In Buenos Aires, the transfer of Telefónica's FTTH network is mandatory.

Personal, the consumer brand of Telecom Argentina, entered 2026 from a position of considerable strength.

At the end of 2025, Telecom reported approximately 19.9 million Personal mobile subscribers in Argentina, excluding the acquired Telefónica business. The subscriber base had declined by 7.8% during the year, largely because inactive prepaid lines with no recorded traffic were disconnected. Around 60% of the remaining customer base was prepaid and 40% postpaid, while mobile ARPU increased by 15.8% in real terms.

Personal has developed into a fully converged communications brand combining mobile, fixed broadband, television, streaming and digital services. Its mobile network has also been a central part of Telecom's technology strategy.

After pioneering Argentina's first DSS-based 5G deployment, Telecom acquired 100 MHz of dedicated 3.5 GHz spectrum in 2023. The company reported 265 5G sites across major cities by the end of 2024 and continued expanding the network during 2025. Personal was subsequently recognised by Ookla as having Argentina's fastest 5G mobile network during 2025, alongside continued recognition for overall mobile network performance.

The challenge for Personal is now much bigger than network performance.

Telecom owns the former Telefónica Argentina business, but regulatory approval requires it to give up six million mobile customers and the spectrum needed to support them. The company must therefore integrate a major acquisition while simultaneously implementing one of the most substantial mobile market remedies seen in the region.

For Operator Watch readers, however, the mobile element is the most intriguing.

• Who will acquire six million customers?
• How much and which spectrum will ultimately accompany them?
• Can the acquiring company build a nationwide network and commercial operation capable of competing with Claro and Personal?
• And how quickly can it move from transitional access arrangements to a sustainable network of its own?

The outcome could reshape not only market shares but also future network investment. A new or strengthened third player could inherit millions of customers from day one, but it would still need to develop its own retail strategy, network footprint, spectrum roadmap, distribution channels and brand proposition.

At the same time, Argentina's operators continue to face economic conditions that make network investment particularly challenging. Telecom equipment and many network inputs are priced in US dollars, while most customer revenues are generated in Argentine pesos. Inflation, currency volatility and access to international financing have all complicated long-term capital planning. Telecom itself identifies inflation, devaluation, exchange-rate risk and the cost of financing capital expenditure among the major risks affecting its operations.

Despite these pressures, investment has continued.

Operators have expanded LTE capacity, fibre backhaul and 5G infrastructure. Personal and Claro each acquired 100 MHz of 3.5 GHz spectrum, while Telefónica acquired 50 MHz before its sale to Telecom. The next phase will be less about simply switching on more 5G sites and more about turning those networks into commercially sustainable platforms.

For consumers, 5G will continue to improve speed and capacity as dedicated mid-band coverage expands. LTE, however, is likely to remain critical for nationwide coverage and the majority of connections for some time.

For operators, the more difficult questions will be around monetisation. These include encouraging migration from prepaid to higher-value contracts, increasing the value of converged mobile and fixed propositions, developing enterprise 5G services and using better network performance to improve customer retention rather than simply competing on price.

Argentina therefore enters the second half of 2026 with two major transitions happening at the same time.

The first is technological, as the market moves from LTE and early DSS-based 5G towards broader use of dedicated 3.5 GHz networks.

The second is structural, as a market that spent years with three established national operators tries to preserve three-way competition after one of those operators was acquired by another.

The regulatory decision does not answer all the questions. In many ways, it creates new ones.

The identity of the buyer of the six million customers will be critical. So will the exact spectrum package, the ability of the new competitor to build its own network and the willingness of investors to finance another nationwide mobile operation.

For years, Argentina's mobile market was defined by competition between Personal, Claro and Movistar.

The next chapter may be defined by Personal, Claro and a third operator that does not yet exist in its final form.

That could make Argentina one of the most interesting mobile markets to watch in Latin America over the next few years.

Telia Shows How 5G Standalone Can Move Critical IoT Beyond Best Effort Connectivity

Telia has announced the launch of Telia Critical IoT Connectivity in Sweden, a new service that uses 5G Standalone technology to support more demanding Internet of Things applications. The announcement is interesting not just because it is another 5G SA milestone, but because it shows how operators are starting to turn the more advanced capabilities of 5G into commercial services for enterprise and public sector customers.

For years, IoT connectivity has often been discussed in terms of coverage, device volumes, battery life and cost. These are still important, especially for massive IoT applications such as smart meters, trackers and simple sensors. But not all IoT is the same. Some applications need much more than occasional data transfer. They need predictable performance, low latency, stronger security and the ability to keep working reliably when the network is under pressure.

That is where Telia’s announcement becomes significant. The company describes Critical IoT Connectivity as a service for highly demanding use cases in sectors such as emergency services, energy, transport, industry and healthcare. These are areas where connectivity is not just useful, but can become part of the operational fabric. A delayed meter reading may be inconvenient, but delayed video from an emergency vehicle, unreliable remote control in an industrial environment, or poor connectivity for critical infrastructure monitoring can have a much bigger impact.

The key technical shift here is 5G Standalone. Many early 5G deployments used 5G Non-Standalone, where the 5G radio was added while the network continued to rely on the 4G core. That helped operators launch 5G services quickly and deliver better mobile broadband, but many of the deeper 5G capabilities depend on the 5G Core. With 5G SA, the radio access network connects directly to the 5G Core, allowing the network to support more advanced service control, security, traffic management and slicing.

Network slicing is one of the most important parts of this story. Instead of treating every connection in the same best-effort way, slicing allows an operator to create logical network slices with different characteristics. One slice could be optimised for high reliability and low latency, another for high-performance video, another for massive IoT, and another for standard best-effort connectivity. The physical network may be shared, but the service behaviour can be differentiated.

For operators, this is an important step. 5G SA is not only about faster speeds for consumers. It is about creating a more flexible platform where different applications can be given different treatment depending on their requirements. That becomes especially relevant for IoT, because a smart meter, a video camera, a connected ambulance, an industrial robot and a remote monitoring system do not all need the same network behaviour.

Telia’s service also highlights the role of the public mobile network in critical IoT. Private 5G networks remain very important, especially for factories, ports, mines, airports, campuses and other controlled environments where local coverage, dedicated infrastructure and tight operational control are required. But there are many IoT use cases that are not limited to a single site. Transport systems, energy assets, emergency services, connected vehicles, field operations and distributed infrastructure often need wide-area connectivity.

This is where public 5G SA with slicing becomes attractive. It can offer a route to more predictable and secure connectivity over a much wider geography than a private network. Telia points to its Swedish 5G network coverage, which reaches 99.9% of the population and more than 94% of Sweden’s land area. That wide-area footprint matters because critical IoT often needs to work across cities, rural areas, transport corridors and distributed operational environments.

The announcement also builds on Telia’s NorthStar 5G innovation programme, which has been used to give partners early access to advanced 5G capabilities. This is worth noting because many of the most interesting enterprise 5G services are unlikely to appear fully formed overnight. They need trials, ecosystem development, operational learning and a better understanding of where customers are willing to pay for differentiated connectivity.

The bigger question for operators is whether 5G SA can help move enterprise IoT beyond simple connectivity pricing. If operators can offer more predictable performance, stronger isolation, security, policy control and service-level differentiation, then critical IoT becomes a more valuable proposition than a standard SIM or data plan. That does not mean every IoT application needs 5G SA or slicing, but it does mean operators have a stronger story for use cases where best effort is not enough.

There is also a need for realism. 5G SA does not automatically make every application ultra-reliable or ultra-low latency everywhere. The actual service experience will depend on how the network is designed, how the slice is configured, the available radio conditions, device capabilities, coverage, service-level agreements and operational processes. But the direction of travel is clear. 5G SA gives operators more tools to create differentiated services, and critical IoT is one of the clearest areas where those tools can be useful.

For the telecoms industry, this is the kind of announcement that matters because it connects technology evolution with commercial service design. 5G SA, 5G Core and network slicing have often been discussed in technical terms. Telia’s Critical IoT Connectivity shows how these capabilities can be packaged into something easier for enterprises and public sector organisations to understand: secure, stable and predictable connectivity for important operational applications.

The real opportunity is not just to connect more things. It is to connect the right things with the right level of performance, reliability, security and control. That is where 5G Standalone could become much more relevant for operators, enterprises and society-critical services.

We have also made a short explainer video on how 5G Standalone is being used for critical IoT, using this Telia announcement as an example:

Related Posts: 

• Private Networks Technology Blog: Private 5G Helps Volvo in Construction, Mining, Quarrying with Autonomous Transportation
• Operator Watch Blog: Sweden is Nordic 5G Laggard
• Operator Watch Blog: Telia 5G Strategy and Services 

Sri Lanka’s Mobile Market Enters the 5G Era After Consolidation

Sri Lanka’s telecoms market has moved through a difficult period since the economic crisis of 2022, when operators had to deal with currency depreciation, higher energy costs, power shortages and weaker consumer spending. The market now appears to be stabilising, but growth is no longer coming from basic mobile adoption. Instead, the story is increasingly about consolidation, mobile broadband usage, spectrum, and the first phase of commercial 5G.

According to data from GSMA Intelligence, by late 2025, Sri Lanka had 30.3 million cellular mobile connections, equivalent to around 130% of the population. That high penetration reflects the familiar pattern of multiple SIM ownership rather than universal individual adoption. At the same time, the total number of connections fell slightly, down by 314,000 or 1.0% during 2025, suggesting a saturated market still shaped by economic pressures.

On the technology side, the market continues to shift towards higher-speed networks. Around 91.3% of all connections are now classified as “broadband”, meaning they operate on 3G, 4G, or 5G networks. However, this should be interpreted with caution: not all of these subscriptions actively use mobile data, as some plans remain limited to voice and SMS services.

The operator landscape has also become simpler. Dialog Axiata completed its acquisition of Airtel Lanka in 2024, reducing the mobile market to three main network operators: Dialog, SLT-Mobitel and Hutch. This consolidation gives Dialog greater scale and spectrum efficiency, but it also increases the importance of regulatory oversight to ensure competition remains healthy.

The most significant change came in December 2025, when TRCSL awarded 5G spectrum following Sri Lanka’s first spectrum auction. Dialog and SLT-Mobitel each received 100 MHz in the 3.5 GHz band, while Dialog also secured 200 MHz in the 27 GHz band for high-capacity 5G deployments. This formally moved Sri Lanka from years of trials and pre-commercial activity into the commercial 5G phase.

Dialog Axiata remains the clear market leader in Sri Lanka, with a subscriber base of over 19 million users, giving it an estimated market share of around 55–60% of mobile connections. The operator has consistently led on technology, being the first to introduce successive generations of mobile networks including commercial 5G services, and continues to expand its next-generation footprint through partnerships such as its 5G RAN deployment with Ericsson and is now leading the country’s transition to 5G, having launched commercial services with over 220 live sites initially serving more than 1.5 million users, and rapidly expanding to over 800 sites nationwide within months of launch. Dialog’s early 5G rollout is underpinned by a strong spectrum position, including holdings in both the 3.5 GHz mid-band for coverage and 27 GHz mmWave spectrum for high-capacity use cases, positioning it well for advanced consumer and enterprise applications . The operator has further strengthened its leadership through the acquisition of Airtel Lanka, enabling greater scale, spectrum efficiency, and network synergies in an increasingly consolidated market.

Hutch, the Sri Lankan operation of CK Hutchison Holdings, remains the smallest of the country’s three mobile network operators but continues to position itself as a value-focused challenger. Hutch serves around 3.6–3.8 million subscribers, giving it an estimated market share of 10–12%, and positioning it as the third-largest provider.

The operator gained scale following its 2018 acquisition and integration of Etisalat Lanka, creating a combined entity with an estimated market share of around 25–26% and establishing it as the third-largest provider, just behind SLT-Mobitel. Hutch has since focused on expanding 4G coverage, having initially launched LTE services in the Western Province in 2018 before extending its footprint across the country, particularly in densely populated areas, while maintaining a strong emphasis on competitively priced data offerings.

On the technology front, Hutch has demonstrated early 5G readiness, conducting its first live 5G trial in March 2021 at its One Galle Face experience centre after receiving a trial licence from the Telecommunications Regulatory Commission of Sri Lanka. The trial, carried out in partnership with ZTE, achieved peak speeds of around 1.8 Gbps alongside low latency performance and showcased use cases including cloud gaming, remote collaboration, and high-definition video streaming. Despite these early demonstrations, Hutch has yet to move to large-scale commercial 5G deployment, reflecting a more measured investment approach compared to larger rivals.

SLT-Mobitel, the mobile arm of Sri Lanka Telecom, has a legacy stretching back over 160 years, originating with the first telegraphic circuit between Galle and Colombo in 1858, and the country’s first international telegraph link to India.. Today, SLT-Mobitel serves an estimated 6–6.5 million subscribers, representing roughly 18–20% of the market, positioning it as the second-largest operator after Dialog.

The operator has expanded its pre-commercial 5G network to key cities including Colombo, Kandy, Anuradhapura, Galle, and Jaffna, offering both mobile broadband and fixed wireless access (FWA) services. In 2021, SLT Group CEO Lalith Seneviratne confirmed plans for a pre-commercial 5G launch using spectrum in the 3.5 GHz band allocated by the Telecommunications Regulatory Commission of Sri Lanka, initially targeting enterprise applications. Full commercial deployment is contingent on final spectrum licensing, with an estimated one-year rollout period once approvals are issued. SLT-Mobitel has earmarked around USD 100 million for 5G network deployment, in addition to approximately USD 15 million for spectrum acquisition.

Sri Lanka’s mobile market has evolved rapidly since the economic turbulence of 2022, moving from a period of consolidation and cautious investment to one of technological advancement and 5G deployment. Dialog Axiata, with its dominant 55–60% market share, continues to lead the market through aggressive 5G rollout, spectrum strength, and the integration of Airtel Lanka, securing its position at the forefront of innovation. SLT-Mobitel leverages its long-standing infrastructure legacy and strategic investment in pre-commercial 5G to maintain its role as the second-largest operator, targeting both urban and enterprise users. Meanwhile, Hutch has carved out a value-focused niche, expanding 4G coverage nationwide and demonstrating early 5G capability through trials in partnership with ZTE, positioning itself as a smaller but agile competitor.

With 5G networks now operational in key cities, and operators investing in LTE densification and spectrum refarming for wider coverage, Sri Lanka is entering a new phase of digital connectivity. While challenges such as device affordability, rural coverage, and ROI considerations remain, the combination of network upgrades, regulatory support, and market consolidation suggests that Sri Lanka is well on its way to becoming a mature, 5G-enabled mobile market, offering both consumers and enterprises faster, more reliable, and innovative mobile services.

Overall, Sri Lanka is entering a more mature phase of mobile market development. Subscriber growth is limited, the number of operators has reduced, and 5G is now moving from demonstrations to commercial service. The key questions are no longer whether Sri Lanka will launch 5G, but how quickly coverage will expand, how affordable 5G devices and tariffs will be, and whether operators can turn new spectrum into profitable consumer and enterprise services.

Related Posts: 

• Operator Watch Blog: Dialog Launches First Immersive Metaverse, Futureverse, in Sri Lanka
• Operator Watch Blog: Sri Lanka Mobile Operators struggling in face of Rising Operational Costs and Inflation 

BT Turns Sovereign Cloud into a Telco Opportunity

BT’s latest sovereign services announcement is more than another cloud launch. It is a sign that digital sovereignty is becoming a serious operator opportunity, especially as AI adoption, national resilience and enterprise risk management begin to converge.

BT Business has announced what it describes as the UK’s first full suite of sovereign services, covering sovereign connectivity, voice, cloud and AI for public and private sector organisations. The announcement is supported by new research from Assembly Research, which suggests that digital sovereignty could unlock an £18 billion productivity opportunity for the UK by giving organisations greater confidence to scale AI securely.

That headline figure is useful, but the more interesting story is what this means for telecoms operators. For many years, telcos have looked for credible ways to move up the value chain beyond connectivity. Sovereign cloud and sovereign AI may provide one such route, not because operators can out-hyperscale the hyperscalers, but because they can bring together secure networks, domestic infrastructure, operational trust, public sector relationships and resilience at national scale.

This was also the theme behind Colin Bannon’s FutureNet World 2026 presentation, The Sovereign Cloud Opportunity for Telcos. The talk is useful because it adds context to BT’s announcement and explains why sovereignty is no longer just about where data is stored. The discussion has moved from data sovereignty to operational sovereignty, and in some cases towards technology sovereignty.

A few years ago, the sovereignty conversation was largely about data privacy, GDPR, encryption, key management and data residency. Those issues still matter, but many customers have now moved on to a harder question: can they keep their business running if something goes wrong? In the FutureNet talk, this was framed as the point where cyber risk, resilience and sovereignty meet. For boards and risk committees, the concern is no longer only whether someone can read sensitive data. It is also whether a critical platform can be switched off, disrupted, restricted, or made unavailable at the worst possible time.

This is where the taxonomy shown in the first slide becomes helpful. At one end of the spectrum sits data sovereignty, with familiar concepts such as data residency, data encryption and metadata control. In the middle is operational sovereignty, which includes tooling, legal entity, operations, certificates, licensing continuity, code escrow and immunity from foreign jurisdiction. At the far end is technology sovereignty, where the ambition is full isolation and technical autonomy.

That distinction matters because many cloud discussions use the word sovereign too casually. A workload hosted in a local data centre is not automatically sovereign in any meaningful sense. The provider, platform, legal entity, operational staff, support model, software supply chain and foreign jurisdiction exposure all matter. In other words, sovereignty is not a single feature. It is a spectrum of control.

BT’s newly expanded portfolio appears to recognise this. The company says its sovereign services are designed to help customers keep sensitive workloads in the UK, meet data residency and regulatory requirements, and adopt AI while maintaining security and operational control. The portfolio spans connectivity, voice, cloud and AI, which is important because sovereignty does not stop at compute. Critical services also depend on networks, communications platforms, monitoring, support, operations and service continuity.

BT had already signalled this direction in December 2025, when it announced a sovereign platform intended to underpin new sovereign voice, cloud and AI services. That earlier announcement positioned the platform as a foundation for UK digital sovereignty, with BT planning to make sovereign options available across a broader set of existing products during the first half of 2026.

The FutureNet talk also made an important engineering point. Cloud resilience is often described in terms of zones and regions, but recent outages have reminded enterprises that multiple zones do not always remove the risk of a common failure plane. For critical workloads, some organisations may increasingly look at active-active architectures across different environments. If they are already considering a second location or second platform for resilience, placing that workload on a sovereign platform may help address another category of risk at the same time.

That is a powerful argument for telcos. Operators are part of national critical infrastructure. They already run complex, highly available networks. They understand regulated environments, emergency services, enterprise SLAs and national infrastructure dependencies. They also have domestic operational footprints, field engineering capability and long-standing relationships with government and large enterprises. That does not automatically make them cloud providers, but it does give them a credible right to play in sovereign platforms.

The slide above, showing BT’s Sovereign Cloud Framework, helps explain how BT is thinking about this market. The framework distinguishes between public cloud, controlled cloud, trusted cloud, domestic cloud and private data centre models. It also separates characteristics such as provider, host and run model, customer environment, sovereignty controls and objectives. Below that, it maps sovereignty parameters including legal entity, platform, data residency and people.

This is a useful way of cutting through the marketing noise. Public cloud with some additional controls may be enough for many workloads. A trusted cloud instance operated through a domestic partner may be suitable for others. More sensitive workloads may require a domestic provider or even a private data centre owned by the customer. There is no single answer, because sovereignty requirements vary by workload, sector, risk appetite and regulatory exposure.

BT’s own Sovereign Cloud proposition is positioned as a private cloud platform hosted and operated in the UK for organisations with sensitive or regulated workloads. The company’s announcement says it is supported by Rackspace Technology’s UK data centre infrastructure, UK-based security-cleared teams and managed services for migration, operations and compliance.

The AI element adds another layer. BT has also announced plans to work with Nscale to deliver sovereign AI data centres in the UK using NVIDIA full stack AI infrastructure. Under the plans, Nscale will build up to 14 megawatts of AI data centre capacity across three existing BT sites, while BT provides the infrastructure and connectivity needed to support the deployment.

This is where sovereign cloud becomes a growth opportunity rather than just a defensive compliance offer. AI adoption is often slowed by concerns around data protection, intellectual property, regulatory exposure, model governance and operational control. If enterprises and public sector bodies can run AI workloads domestically, with clearer control over where data is stored, who can access it and how services are operated, then more sensitive use cases become easier to justify.

For operators, this is not about pretending that they can build an entire technology stack from scratch. The FutureNet talk was realistic on this point. Full technology sovereignty is extremely difficult, expensive and, in many cases, undesirable. Modern infrastructure depends on global hardware, software and cloud ecosystems. Cutting off major innovation engines would not make economic or technical sense for most telcos.

The more realistic question is whether a telco can design, build and operate a platform domestically while still working with global partners. In BT’s case, the argument is not about rejecting hyperscalers, NVIDIA, Rackspace, Nscale or other partners. It is about creating a sovereign operating model around the parts that matter most to customers: location, control, legal exposure, operations, resilience and trust.

There is still a major challenge. Sovereignty remains poorly defined. The FutureNet presentation referred to the risk of sovereign washing, where products are labelled sovereign without a clear explanation of what that actually means. This creates problems for customers and providers alike. If one provider invests in domestic operations, security-cleared staff and locally controlled platforms, while another simply adds a sovereignty label to an existing service, customers need a way to compare them fairly.

This is where regulation or at least an industry framework may become important. The UK has so far taken a relatively open market approach, which encourages innovation but can also create uncertainty. Without common definitions, it is hard for customers to know what they are buying and hard for providers to justify major investments. A shared framework for levels of sovereignty could reduce confusion, support investment and make procurement more transparent.

The commercial question is also important. Sovereign platforms can cost more to build and operate, especially if staff, support, infrastructure and operational controls are brought onshore. Customers may value sovereignty, but not every workload will justify a premium. The market will therefore need clearer segmentation. Some workloads may only need data residency. Others may require operational sovereignty. A smaller set may require highly isolated environments.

This is why BT’s full-suite approach is interesting. By combining sovereign connectivity, voice, cloud and AI, BT is trying to present sovereignty as an end-to-end service model rather than a single cloud product. That fits the way critical organisations actually operate. They do not just need compute. They need networks, communications, security, continuity, support, compliance and governance.

For the Operator Watch audience, the bigger takeaway is that sovereignty could become one of the more credible routes for telcos to participate in cloud and AI. The opportunity is not to become generic public cloud providers. That battle has already been won by others. The opportunity is to become trusted national infrastructure partners that help enterprises and government bodies run critical workloads with greater assurance.

BT’s announcement is therefore both a product story and a positioning story. It shows how an incumbent operator can connect its network heritage, public sector role, security capabilities, cloud partnerships and AI infrastructure ambitions into a single narrative. Whether customers buy into it at scale will depend on pricing, service clarity, workload suitability and the development of common market definitions.

What is clear is that sovereignty is moving from the margins to the centre of the enterprise technology conversation. As AI becomes more embedded in critical workflows, organisations will increasingly ask not only what the technology can do, but where it runs, who controls it, who can access it and whether it will still be available on a bad day.

That is a conversation where telcos have a stronger role than they have had in many previous cloud debates. BT is now making that case explicitly. The success of this strategy will depend on whether the market sees sovereign services not as a compliance overhead, but as a foundation for trusted AI, resilient operations and national digital capability.

The video of the presentation as follows:

Related Posts: 

• The 3G4G Blog: IET Lecture by Prof. Andy Sutton: Point to Point Microwave Radio Systems
• The 3G4G Blog - Towers, Masts and Poles: The Backbone of Telecom Infrastructure
• Telecoms Infrastructure Blog: BT/EE’s Growing Dependence on Small Cells to Boost Network Performance
• Operator Watch Blog: EE Combines More 4G and 5G Spectrum for Even Higher Data Rates
• Connectivity Technology Blog - 5G Demystified; the What, When and Where by Prof. Andy Sutton, BT
• Operator Watch Blog: BT UK's Journey to Automation
• Operator Watch Blog - EE: 5G is going to change everything all over again 

Serbia Enters the 5G Era as Operators Accelerate Nationwide Upgrades

Serbia’s mobile telecommunications market has evolved significantly over the past few years. While the country spent several years preparing for next-generation connectivity through trials and infrastructure upgrades, commercial 5G services only became a reality at the end of 2025.

Data from GSMA Intelligence indicates that Serbia had 8.45 million cellular mobile connections at the end of 2025.

To provide context, it is common for the number of mobile connections to exceed the total population because many individuals use more than one SIM or mobile subscription. For instance, a person may have one connection for personal use and another for work purposes. The growing availability of eSIM technology has made it even easier for users to maintain multiple mobile connections in recent years.

GSMA Intelligence data shows that the total number of mobile connections in Serbia was equivalent to 127 percent of the country’s population in October 2025. Looking at recent trends, the total number of mobile connections in Serbia declined by around 215 thousand, or 2.5 percent, between the end of 2024 and the end of 2025. The data also indicates that 95.4 percent of mobile connections in Serbia can now be classified as broadband connections, meaning that they operate on 3G, 4G, or 5G mobile networks.

However, it is important to note that devices capable of connecting to broadband mobile networks do not necessarily use cellular mobile data. Some subscriptions, for example, may only provide access to voice calls and SMS services. As a result, the share of broadband connections should not be interpreted as a direct measure of mobile internet usage.

Today, the Serbian mobile market is served by three nationwide operators: Telekom Srbija (mts), Yettel Serbia (formerly Telenor Serbia), and A1 Serbia (formerly Vip mobile). Together they cover the entire country and are now transitioning from mature LTE networks toward full 5G deployments.

Based on July 2025 data from Opensignal, Yettel leads the Serbian mobile network market with the best overall experience, securing multiple performance and coverage awards. A1 is a strong competitor, winning several categories, while all operators offer 2G, 3G, and 4G, with 5G deployment expanding in major cities like Belgrade, Niš, and Novi Sad.

Telekom Srbija, operating under the MTS brand, remains the largest telecom operator in the country. It is majority state-owned and offers mobile, fixed broadband, television, and enterprise services.

The operator holds the largest share of the Serbian mobile market, accounting for about 42.7% of the country’s 8.1 million mobile subscribers as of 2024.

Telekom Srbija also leads in network infrastructure. According to regulator reports, it had the highest number of base stations across 2G, 3G, and 4G networks by the end of 2024, reflecting its strong nationwide footprint.

The company had previously conducted 5G trials with vendors such as Ericsson, testing use cases and performance in urban areas. These trials laid the groundwork for commercial deployment once spectrum licences were issued.

The second largest operator is Yettel Serbia, which was previously known as Telenor Serbia before a rebranding in 2022. The company is now part of the e& PPF Telecom Group and provides mobile, broadband, and digital services nationwide.

Yettel holds roughly 33% of Serbia’s mobile market, making it the second-largest operator behind Telekom Srbija.

In preparation for 5G, the operator invested more than €100 million to secure spectrum across key bands such as 700 MHz, 2600 MHz and 3.6 GHz. These frequencies enable a mix of wide-area coverage and high-capacity services required for next-generation networks.

The company has been positioning itself as a strong competitor through network upgrades and digital services, including fixed wireless broadband solutions based on 5G technology.

The third operator is A1 Serbia, part of the A1 Telekom Austria Group. Although smaller than its rivals, it remains a significant player in the Serbian telecom market.

As of 2024, A1 Serbia accounted for around 24.3% of mobile subscribers.

A1 has focused heavily on network modernization and digital infrastructure investments. The company has also announced major investment programmes aimed at expanding fibre infrastructure and strengthening its mobile network in preparation for the 5G era.

Serbia’s long-awaited transition to 5G finally took place in late 2025. The national regulator, RATEL, conducted a multi-band spectrum auction in November 2025 covering frequencies including 700 MHz, 900 MHz, 1800 MHz, 2100 MHz, 2600 MHz and 3.6 GHz.

All three operators: Telekom Srbija, Yettel, and A1, secured licences in the auction, each paying just over €100 million, bringing the total value of the spectrum allocation to around €300 million.

Shortly afterwards, the operators simultaneously activated commercial 5G networks in December 2025, marking the country’s official entry into the 5G era.

Initial coverage focuses on major cities and key tourist areas, with broader nationwide rollout planned through 2026. Users with compatible devices can access the new network where coverage is available.

Serbia’s telecom sector is now entering a new phase of development as operators shift their focus from LTE expansion toward advanced 5G services.

In the short term, operators are expected to concentrate on expanding coverage in urban areas and introducing 5G-based fixed wireless broadband as an alternative to legacy fixed connections.

Over the longer term, the deployment of standalone 5G cores and the expansion of fibre infrastructure will likely play a key role in enabling new digital services, supporting industry modernization, and improving broadband connectivity across both urban and rural regions.

Serbia’s entry into the 5G era represents a significant milestone for the country’s digital economy. With spectrum secured, infrastructure investment under way and competitive pressure among operators increasing, the market is positioned for substantial progress in the years ahead.

Vodafone Approach to AI Driven Networks Built on APIs

At FutureNet World 2026 in London, one of the more thought-provoking sessions came from Víctor Fernández of Vodafone Group, who explored what it really means to enable AI to act at scale through APIs. While much of the industry conversation around AI still focuses on models and data, this talk shifted attention to a more practical layer. If AI agents are to move beyond insights and into action, they need structured, reliable and well-designed interfaces into the network and its capabilities.

The central idea was straightforward but important. AI agents cannot deliver value in isolation. They need access to systems, data and control points, and APIs are the mechanism that makes this possible. However, simply exposing APIs is not enough. If they are not designed with AI consumption in mind, they risk becoming bottlenecks rather than enablers.

A key theme was the need to rethink API design principles for an AI-driven environment. Traditional APIs have largely been built for human developers who understand documentation, constraints and context. AI agents operate differently. They require APIs that are easy to discover, clearly defined and consistent in structure. Standardisation becomes critical here, not just for interoperability across systems, but also to allow agents to reliably interpret and use them without ambiguity.

Closely linked to this is composability. AI agents are most powerful when they can orchestrate multiple services to achieve a goal rather than executing a single isolated task. This means APIs should not be designed as standalone endpoints, but as building blocks that can be combined into workflows. In a telecoms context, this could involve stitching together network data, policy controls and service exposure capabilities to enable more advanced automation scenarios.

Another important aspect discussed was scale. As AI agents begin to interact with APIs at a much higher frequency than human users, the load on systems will increase significantly. This introduces new challenges around monitoring, management and optimisation. It is no longer just about whether an API works, but how it behaves under continuous, automated usage. Observability, rate control and intelligent traffic management become essential to ensure reliability.

There is also a governance dimension. Allowing AI agents to take actions through APIs introduces risks if not properly controlled. Mechanisms are needed to define what actions are permitted, under what conditions, and with what level of oversight. This is particularly relevant in telecom networks where certain actions can have wide-ranging operational impacts. Ensuring that AI-driven interactions remain safe and predictable is as important as enabling them in the first place.

The session also highlighted the importance of a developer-friendly ecosystem. Even in an AI-centric world, developers remain central to creating and maintaining the underlying capabilities. Clear documentation, sandbox environments and consistent tooling are all necessary to support both human developers and the AI systems they build. In many ways, the goal is to create an environment where APIs are equally usable by humans and machines.

What stood out from this talk was how it grounded the AI discussion in something tangible. There is a growing recognition that the real challenge is not just building smarter models, but integrating them into operational systems in a way that delivers measurable outcomes. APIs sit right at that intersection.

For operators, this represents both an opportunity and a responsibility. By exposing network capabilities through well-designed APIs, they can enable a new class of applications and services driven by AI. At the same time, they need to ensure that these interfaces are robust, scalable and secure enough to handle the demands of automated consumption.

As the industry continues to explore the role of AI in telecoms, conversations like this help move the narrative forward. They shift the focus from what AI could do in theory to what it needs in practice to operate effectively at scale.

The video of the talk is embedded below:

Related Posts: 

• Operator Watch Blog: Vodafone UK Bares Its Vision and European Strategy
• Telecoms Infrastructure Blog: Vodafone Explains Mobile Phone Mast
• Operator Watch Blog: Vodafone to keep pushing for Open RAN
• Operator Watch Blog: Vodafone UK Selects Commercial Open RAN Network Partners
• Telecoms Infrastructure Blog: Vodafone UK's 5G Infrastructure