Gayan Koralage, Director of Strategy at EDOTCO Group, explores how telecom infrastructure is evolving from fragmented connectivity assets into integrated digital platforms that combine connectivity, compute and resilience — creating new opportunities for operators to capture value in the AI-driven economy.

Few weeks ago, Chinese operators including China Mobile and China Telecom quietly introduced AI mobile plans linked to “tokens” instead of only gigabytes. Early customer reactions were mixed, some consumers found it innovative while many struggled to understand what “AI tokens” actually meant. Strategically however, this may signal the next major transition after the mobile internet era.

For the last 30 years, telecom networks were built mainly for human downstream consumption, voice calls, messaging and later video streaming. More than 800 mobile operators globally spent over US$2 trillion in the last decade building spectrum, fiber, towers, subsea cables and 5G infrastructure to carry exploding internet traffic. 

Average smartphone usage globally grew from around 2GB per month five years ago to more than 20GB today in advanced markets, with some countries already exceeding 30GB monthly per user.

The great value shift

Despite this enormous investment cycle, telecom industry revenues remained relatively flat at around US$1.1 trillion annually. Meanwhile, five major technology companies, Apple, Microsoft, Amazon, Alphabet and Meta, increased their combined market capitalization from roughly US$2 trillion a decade ago to more than US$15 trillion today.

Telecom operators tried to challenge this imbalance through the “fair share” debate, arguing large OTT traffic generators like Netflix, YouTube and Meta should contribute towards network investments. Europe debated this heavily, with similar discussions emerging across South America and Australia. Ultimately, the model failed because regulators defended open internet principles, customers demanded unrestricted access and OTT players controlled the digital experience layer while telecom operators remained largely connectivity providers.

AI changes the equation

Over the next five years from 2026 to 2031, global mobile data traffic may triple again as AI-native devices, AI assistants, robots and autonomous systems scale globally. But unlike the previous internet wave denominated by gigabytes, the AI economy may increasingly be denominated by “tokens”, units of intelligence processing.

Hyperscalers and AI companies are now investing at extraordinary scale, potentially more than US$1 trillion globally into GPUs, AI datacentres, subsea cables, AI chips and large language models across Asia, Europe and the United States. Microsoft, Amazon, Google, Meta, Nvidia, OpenAI and sovereign AI initiatives are building what may become the largest infrastructure cycle since cloud computing itself.

AI-native economies ahead

This AI transition is much larger than technology alone. Every government, institution, enterprise, SME and individual may go through an “AI-native” transformation over the next decade. In the past we digitized manual workflows. In the future, workflows themselves become intelligent through AI orchestration.

For example, future national early warning systems may combine rainfall data, river sensor readings, traffic systems and satellite imagery into AI platforms that predict exact flood locations and evacuation timing in real time. Hospitals may use AI to predict diseases before symptoms appear. Factories may operate with autonomous robots and self-healing production lines. The possibilities for humanity are immense.

Telecom’s opportunity

This creates both opportunity and risk for telecom operators.

In the gigabyte era, telecom operators built most of the bandwidth capacity while OTT players captured most of the value. In the AI era, hyperscalers may again invest heavily into backend AI infrastructure. But for the first time, telecom operators may have an opportunity to participate more directly in monetisation because they still control the “customer edge”, the last mile connectivity, subscriber identity, billing relationships, distributed infrastructure and network orchestration.

For the first time, the industry may potentially turn the tables. OTT and hyperscale players may build the AI capacity while mobile operators monetize intelligently at the access point.

How telecom could price AI

The industry may potentially evolve several AI pricing models. Some operators may bundle AI copilots, AI search and productivity assistants into premium mobile plans similar to streaming bundles today. Others may introduce “AI capability tiers” where higher-priced plans unlock more advanced AI services such as multimodal assistants, autonomous agents or video generation.

Enterprise pricing may become even more important. Telecom operators could monetise AI through APIs, enterprise AI subscriptions, low-latency AI connectivity and edge inferencing services for factories, ports, robotics and smart cities. Future pricing models may increasingly combine bandwidth, latency and compute together instead of selling only gigabytes.

However, telecom operators may need to work collectively rather than individually. If operators remain fragmented while hyperscalers scale globally, the industry risks repeating the OTT era again. This may require GSMA-led frameworks around network APIs, AI orchestration standards, edge interoperability and common charging principles so operators can collectively participate in the AI value chain instead of competing only on connectivity pricing.

The critical lesson from the OTT era

The OTT era taught the telecom industry an important lesson, connectivity alone is no longer enough. The next AI wave may therefore require operators to think and act differently, not individually but collectively.

Rather than competing only on bandwidth pricing, telecom operators now have an opportunity to collaborate through GSMA-led initiatives, shared API frameworks, edge orchestration platforms and interoperable AI ecosystems. By working together, operators may evolve from pure connectivity providers into intelligent digital infrastructure partners powering the global AI economy.

This transition will not be without challenges. Token-based models, AI regulation and hyperscaler dominance will continue to evolve rapidly. But unlike previous technology cycles, telecom operators now hold a unique strategic advantage — proximity to the customer edge, distributed infrastructure and real-time network intelligence.

A brave new era ahead

For the first time in decades, telecom may have an opportunity to move higher into the digital value chain rather than remaining only a utility layer. Future telecom networks may not simply move packets faster; they may increasingly move intelligence itself.

If the industry can act fast, think boldly and collaborate collectively, a new era of growth, relevance and innovation may emerge for telecom globally. 

As Alexander Graham Bell once said: “When one door closes, another opens.” Perhaps the AI era is that next door opening for telecom.

Gayan Koralage is Director Strategy and Malaysia Business at EDOTCO Group, one of Asia’s largest neutral-host telecom tower companies. He has been with EDOTCO since its inception in 2013 and has played a key role in its regional expansion across multiple Asian markets. A regular speaker and writer on telecommunications and digital infrastructure opportunities, his work focuses on 5G economics, network transformation, and the role of connectivity in the digital economy.