MWC 2026: To monetise AI, telcos must sell enterprise outcomes

Embedding AI in operational workflows

At the event, discussions moved away from potential and possibilities toward pragmatic matters such as latency, failure tolerance, governance and data sovereignty.

This reflects a broader enterprise reality: AI that sits outside core workflows remains discretionary spend. Conversely, AI that directly improves uptime, throughput, yield or safety competes much more effectively for the budgets owned by CIOs, chief operating officers and line-of-business leaders.

The implication is straightforward – revenue follows deployment, and deployment follows outcomes. The most credible AI narratives at MWC were those demonstrating operational impact rather than technological novelty.

The vertical battleground: The operational AI stack

What became increasingly clear in Barcelona is that enterprise AI is no longer being deployed as a standalone capability. Rather, it combines digital twins, edge AI and dedicated connectivity to form an operational AI stack. It is one system, not three separate projects.

From Singapore to Shanghai, and Beijing to Busan, ports and factories are already running on these AI stacks. These are not proofs of concept; they are real-life operations in commercial use, where digital twins are increasingly used to simulate and optimise physical processes in near-real time.

Meanwhile, edge AI handles time-critical inference for quality inspection and autonomous vehicles, while private or campus networks provide deterministic latency, security and availability. These elements are interdependent, and enterprises do not purchase them piecemeal.

This interdependency explains why vertical AI is proving monetisable. Enterprises are not buying AI per se. They are buying higher asset utilisation, faster throughput and safer factories. Crucially, this AI stack requires telecom capabilities – latency, resilience and regulated data paths – which gives operators pricing power when bundled as managed outcomes.

The road to revenue

For telecom operators, this creates a narrow but credible path to AI revenue. Operators tend to struggle when competing with hyperscalers on foundation models or generic platforms. Instead, value capture lies at the intersection of AI, connectivity and industry operations.

Orange CEO Christel Heydemann articulated this well at MWC. She argued that operators must evolve from connectivity providers to architects of trust, turning resilience, cyber security and data protection into competitive advantages. At the same time, she acknowledged the economic imbalance facing operators, noting that despite operating critical digital infrastructure, telcos are not reaping rewards proportionate to their role.

Trust and performance are not abstract concepts when placed in the context of an operational workflow within enterprise verticals. When AI stacks control physical processes, operators can price predictability, accountability and compliance. Multi-year managed services around private 5G, edge compute and secure data pipelines can then be bundled and contracted out to enterprises.

APAC shows what execution looks like

Asian operators at the event provided a blueprint for how to monetise AI. SK Telecom CEO Jung Jaihun stated that operators’ proprietary infrastructure and operational expertise are key to building AI infrastructure, adding that telcos must move beyond data delivery to play a leading role in shaping AI services for enterprises.

Singtel made a similar argument, but with a strong emphasis on outcomes. Announcing its expanded 5G Advanced strategy at MWC, Singtel Singapore CEO Ng Tian Chong said: “We see 5G Advanced not as a network upgrade, but as a foundational platform for Singapore’s AI-powered future. By embedding intelligence, programmability and service differentiation into our network, we enable enterprises to innovate with confidence and compete globally.”

This framing from SK Telecom and Singtel is significant. They are not selling AI as a standalone product or an overlay; rather, AI is inseparable from infrastructure and industry services. This integrated stack combines private networks, edge compute and service-level agreements (SLAs). In sectors such as manufacturing and logistics, this bundling approach aligns closely with how enterprises buy technology.

Why execution will decide AI winners

Many operators understand where AI value is forming, but few are organisationally prepared to capitalise on it. Vertical AI requires a different path to market, new ecosystem partnerships, and a willingness to move away from selling bandwidth to selling outcomes.

The battleground for operators is clear: it is not in the large language model (LLM) or raw compute power, but in the enterprise vertical. Operators that align AI with operational outcomes, trust and performance will find real revenue. Those that do not will continue to enable AI growth, but without getting their fair share of the rewards.

Edwin Lin is principal consultant at Omdia, part of Informa TechTarget