Telcos built their business on selling bandwidth. That business model is running out of road. Across seven sessions at ATxEnterprise 2026, operators, analysts and infrastructure vendors converged on the same conclusion: 5G connectivity alone no longer justifies the capital telcos are pouring into their networks, and AI-powered telcos are the only ones with a credible path to the next decade of revenue growth. The future of connectivity in Asia is being rebuilt around three converging layers — 5G and Open RAN, satellite and hybrid networks, and AI — and the operators moving fastest on all three at once are the ones pulling ahead.

The numbers explain the urgency. According to GSMA's Mobile Economy Asia Pacific 2026 report, mobile technologies and services are on track to contribute US$1.4 trillion to the region's economy by 2030 — a 40% jump from roughly US$1 trillion today — with operators committing more than US$200 billion in capital expenditure between 2025 and 2030 to get there. Yet 5G connections in the region are only expected to reach 50% of all mobile connections by 2030, and Southeast Asia specifically is projected to hit just 43% 5G adoption by that date, trailing well behind developed Asia Pacific markets. The infrastructure for better 5G connectivity is being built. The revenue model to justify it is still being figured out — and that's exactly the gap this year's sessions on connectivity in Asia tried to close.

That gap looks different depending on where you sit within connectivity in Asia. A market like Singapore, with near-universal coverage and a sophisticated enterprise base, is asking how to extract more value per connection. A market like Pakistan or Mauritius is still asking how to extend basic coverage. Both conversations played out across ATxEnterprise 2026's telco track, and both point to the same underlying shift: connectivity in Asia is no longer a single story about coverage — it's a set of parallel stories about monetisation, resilience and AI, each moving at its own speed.

5G connectivity is mature for consumers, unproven for enterprise

The "5G and Future Connectivity" session was blunt about where things stand: consumer-grade 5G connectivity is mature, but enterprise 5G connectivity — the smart manufacturing, mission-critical communications and logistics use cases that were supposed to fund the network — is still experimental. Panellists including Dr David Soldani of Rakuten Mobile pointed to private 5G networks as the most viable near-term path to enterprise monetisation, provided operators can standardise deployment models instead of building bespoke 5G connectivity for every customer. Dr Zhang Kai of China Unicom Singapore made a similar point from within Singapore's telecom sector: consumer and enterprise 5G connectivity now require genuinely different commercial models, not the same product sold twice.

This is where Open RAN enters the picture. Omdia's 2026 Open RAN Operator Survey found that 40% of operators have now meaningfully adopted Open RAN — either running it in production or using its principles to guide their RAN evolution — and that this cohort is markedly more confident about layering AI-driven automation and agentic RAN frameworks on top of their networks. Rakuten Mobile's own numbers, discussed at the "Open RAN: AI, Cloud and Seamless Integration" session, made the case concrete: more than 60,000 base stations deployed, full ownership of the network stack from OSS to BSS to AI applications, and a 40% reduction in operational costs as a result of moving to Open RAN. Michael Tadault of Red Hat framed this as the real inflection point for AI-powered telcos — a shift from vendor silos to open telco cloud models that reduce single-vendor dependency, even as many telcos still hesitate to adopt Open RAN fully because of integration complexity.

Hybrid networks and satellite connectivity are becoming core infrastructure, not backup

The second infrastructure theme was the rise of hybrid networks. "Integrated Connectivity: Building Seamless Hybrid Networks" made the point directly: satellite connectivity, once treated as a fallback for remote or disaster-hit areas, is now a core layer of the network stack, not a supplement to it. Speakers from Eutelsat, Azercosmos and K2 Space described dynamic orchestration inside modern hybrid networks that lets traffic switch seamlessly between terrestrial, mobile and satellite paths without the user ever noticing — a capability increasingly required by enterprise and government customers who need uninterrupted service regardless of what's happening on the ground.

That shift is backed by where the industry standards are heading. As Ericsson's technology research on satellite direct-to-device communication lays out, 3GPP formalised satellite support for 5G with New Radio Non-Terrestrial Networks (NR-NTN) in Release 17, giving operators a standards-based path — rather than a patchwork of proprietary systems — to blend satellite connectivity and terrestrial coverage into genuinely unified hybrid networks. The "Fireside Chat: Future of Connectivity in an Age of AI, 5G and Satellite" session, featuring Space42's Jassem Nasser, tied this convergence directly to 3GPP standardisation and the demand for seamless coverage across 5G and emerging 6G networks. The same session flagged interoperability, data sovereignty and spectrum harmonisation as the unresolved friction points in scaling hybrid networks across borders — echoing calls from the 5G session for predictable, "use it or lose it" spectrum policy to stop hoarding and unlock better satellite connectivity and terrestrial coverage together.

AI is where telcos are trying to find the money

If 5G, Open RAN and satellite are the infrastructure layer, AI is where AI-powered telcos are betting they'll finally make the network pay. CRN Asia's reporting on the region's telco strategy captures the shift in one line: operators across the region are "no longer content to sell connectivity and lease capacity to hyperscalers," and are instead building positions across compute, cloud orchestration, managed security and packaged enterprise AI. "From Connectivity to Intelligence: How Telcos Move Up the Stack to Power AI in the Real World" argued that telcos hold a structural advantage few other industries have: towers, central offices and fibre networks that can be converted into edge AI infrastructure. Giuseppe Ferraioli of Armada made the case that this positions AI-powered telcos to serve latency-sensitive use cases — real-time monitoring, disaster response, defence and border security — that centralised cloud simply can't reach fast enough, with Microsoft's investment in Armada cited as evidence that hyperscalers see telco edge infrastructure as complementary rather than competitive.

The "C-suite Vision: Monetising Intelligence and Connectivity" panel put real operator strategy behind this idea. Keith Leong of Singtel, speaking for Singapore's telecom sector, described an ambition to become an AI-first telco, using network slicing and AI-driven operations to simplify service delivery for both enterprise and consumer customers — a clear signal of where Singapore's telecom sector is placing its bets. Mauritius Telecom detailed investment in localised, Creole-specific AI models as a differentiation strategy few global players can replicate, while Pakistan Telecommunications Company Limited is still focused on foundational 5G connectivity buildout for an underconnected population — a reminder that AI-powered telcos across this region are moving at genuinely different speeds. What every panellist agreed on, though, was that AI's direct monetisation as a standalone product is still uncertain — the near-term value for AI-powered telcos is in embedding AI into existing operations to cut costs and improve customer retention, not in selling "AI" as a line item.

"Digital Telcos: Transforming Connectivity in a Hyperdigital World" reinforced this from the customer-experience side, and made the case that AI-powered telcos win or lose on execution, not ambition. Kelvin Chua of Circles.Life and Jensen Ooi of Omdia argued that frameworks like VIMM (reducing visual, intellectual, memory and motor overload) and CREED (consistency, relevance, efficiency, empathy, delight) matter more than the underlying technology — instant, frictionless 5G connectivity when a customer lands in a new country, they said, will do more for loyalty than any AI feature that isn't solving a real problem.

What this means for connectivity in Asia and Singapore's telecom sector

Put together, these seven sessions describe a telco stack being rebuilt in three layers at once: 5G connectivity and Open RAN at the radio access layer, hybrid networks and satellite connectivity filling in coverage and resilience, and AI sitting on top as the monetisation and orchestration layer that's supposed to make the other two pay for themselves. That's a heavier lift for AI-powered telcos operating across connectivity in Asia than the framing suggests — GSMA's own data shows more than 700 million adults in the region remain offline despite living under mobile broadband coverage, and Southeast Asia's 5G connectivity rollout still lags well behind developed markets.

For Singapore's telecom sector specifically, where enterprise customers and government agencies are pushing hard on data sovereignty, resilience and AI-first service models, the sessions suggest the operators moving fastest are the ones treating Open RAN, satellite connectivity and AI not as three separate initiatives but as one converged telco business model. Singtel's AI-first ambitions, discussed at length in the C-suite Vision panel, are as much a signal about the direction of Singapore's telecom sector as they are a single-company strategy — and the wider region's operators, from Mauritius Telecom to PTCL, are watching what Singapore's telecom sector does next as a bellwether for their own telco business models.

Singapore's telecom sector also has a structural advantage the sessions kept returning to: it sits at the centre of the region's connectivity in Asia conversation, hosting the O-RAN Alliance activity, the robotics precinct referenced in the Open RAN session, and a density of hyperscaler and satellite-operator partnerships that smaller markets in connectivity in Asia simply can't match. That makes Singapore's telecom sector as much a proving ground for AI-powered telcos as it is a market in its own right — decisions made here about Open RAN, satellite connectivity and 5G connectivity tend to set the template that the rest of connectivity in Asia follows.

The throughline across every session was the same: telcos that keep treating 5G connectivity, satellite connectivity and AI as bolt-ons will keep struggling to monetise them. The ones rebuilding their telco business models around an open, AI-orchestrated, satellite-integrated stack — the approach Rakuten Mobile, Singtel and Space42 all described in different ways — are the AI-powered telcos positioned to capture the next wave of enterprise and government demand across connectivity in Asia.

Whether your organisation sits inside Singapore's telecom sector or is watching it from elsewhere in the region, the direction of travel is the same. Want the full breakdown of every session at ATxEnterprise 2026, including the panels behind these takeaways? Download the ATxEnterprise 2026 post-show report for the complete data, or browse the full AI-generated session summary library to revisit any conversation in detail. And if 2027's telco stack is where your organisation is headed next, pre-register for ATxEnterprise 2027 to be in the room for what comes after.

This article draws on official AI-generated session summaries from ATxEnterprise 2026, Asia Tech x Singapore's enterprise tech conference held at Singapore EXPO.