Industry
Communications, Media and Information Services
Connect. Curate. Comprehend: Unlocking Value Beyond Boundaries
Highlights
- The traditional telecoms model is evolving, with operators exploring new growth opportunities as returns moderate and cost structures transform, driving a stronger focus on strategies beyond connectivity sales.
- Distributed AI grids, combined with tokenized network capabilities, offer the potential to monetize different elements of the telco infrastructure, such as latency and sovereignty, via programmable marketplaces. The greatest value, however, lies in enabling enterprises to orchestrate activity over the AI grid.
- Telcos are uniquely placed to lead in this emerging world of AI grids and tokenized services. Their deep experience in managing distributed networks and their role as trusted providers of critical national infrastructure provide a strong competitive advantage.
On This Page
On This Pageinpage
Telcos at a pivotal moment
The telecoms business model is at an inflection point.
Network activity is increasing, while revenues are stabilizing amid more competitive, commoditised pricing. Operating costs are evolving in response to a dynamic geopolitical climate, with energy prices remaining a key concern. At the same time, telco leaders are focused on realizing value from 5G and edge computing investments, while shaping priorities for future growth.
Against this backdrop, it’s vital for telecoms operators to explore new growth models when determining their future building on past successes while adapting to emerging opportunities for value creation.
AI grids together with tokenized network capabilities offer a potential shift in approach, helping telcos move from selling commoditised bandwidth to monetizing intelligent services over their networks.
Understanding AI grids
An AI grid is a geographically distributed, integrated computing fabric for AI that spans the core data center, cloud, and edge zones.
The term ‘AI grid’ has only recently gained prominence as vendors, have adopted it to describe integrated AI execution across multiple computing environments. While the terminology may be relatively new, many of the underlying concepts are well established, including:
- AI infrastructure: the technology foundation required to build, run, manage and scale AI workloads.
- Edge computing / AI: distributed computing or AI frameworks where data is processed in physical computing environments close to the data source, rather than sent for centralized processing.
- Distributed inference: executing AI model requests across multiple computing environments rather than a central location.
As long-established operators of distributed infrastructure, telcos are well positioned to play a leading role in the emerging AI grid ecosystem. More importantly, AI grids give operators the opportunity to expand their offerings beyond connectivity and evolve their networks into programmable platforms that deliver value-added services to enterprise customers.
Rethinking AI economics
The draw of AI grids for enterprises is that they enable users to run AI workloads wherever it makes the most operational and economic sense.
This is particularly appealing in light of the AI pricing reset underway across many AI vendors. Until recently, the prevailing approach was to charge a fixed price per seat, providing stability and visibility over future costs.
Now, AI companies are shifting to a consumption-based pricing model, introducing a greater flexibility in how enterprises manage and align with costs with usage.
Consequently, CTOs and CIOs must balance plans to scale AI in their enterprises with the need to optimize associated costs, heightening the appeal of a distributed approach that reduces reliance on running all AI usage through premium centralized cloud environments.
The commercial layer
Tokenized network capabilities deliver the commercial layer of AI grids.
By converting different capabilities within the AI grid, such as computing power, latency and sovereignty, into measurable units (“digital tokens”), telcos can create more flexible commercial models. So, rather than monetizing bandwidth alone, operators can make different infrastructure capabilities available to customers by packaging them as programmable services that are discovered, consumed and charged via a marketplace.
This approach enables telcos to monetise and differentiate their service offerings more effectively, while also allowing enterprise customers to benefit from a wider range of connectivity-related products and granular, usage-based pricing aligned with their AI requirements.
Trusted data layer
Tokenized network capabilities can also support automated data exchange and collaboration across AI grids.
This is particularly valuable for enterprises involved in multi-party implementations, where AI workloads must be coordinated across a complex AI ecosystem and a varied computing environment.
A real-world example of this scenario was piloted by TCS and AWS with a North American telco. It details how distributed inference and AI agents were deployed to help manage traffic safety at busy urban intersections, using a variety of computing environments to process and analyse data from continuous roadside camera feeds. These included roadside units for far-edge processing, the telco network for near-edge processing and connectivity, and the AWS regional data centre for cloud processing.
Telcos are well positioned to provide a trusted data layer in scenarios like these. In most markets, they are considered critical national infrastructure and are subject to robust security requirements, making them an ideal partner for managing secure data exchange and ensuring compliance with increasingly complex digital sovereignty requirements.
In most markets, they are considered critical national infrastructure and operate under robust security frameworks, making them a strong partner for managing secure data exchange and supporting compliance with evolving digital sovereignty requirements.
AI control planes
Creating a marketplace and facilitating secure data exchange is only part of the story.
Orchestration of AI workloads across the grid and token optimization will become increasingly important, especially as AI ecosystems grow more complex. This is where the control plane becomes increasingly important.
Historically, telecom control planes determined how traffic flowed across the network and how resources were allocated. In an AI grid environment, that role expands beyond connectivity.
AI increasingly becomes part of the control plane itself. It helps determine where inference should run, select the appropriate compute environment, apply trust and sovereignty policies, and optimize cost and performance in real time. In this model, networks don’t just transport the information; they actively coordinate the creation and delivery of intelligence.
To help operators monetize and scale in this new AI paradigm, TCS delivers Industry aligned scalable AI orchestration capabilities that help design , deploy , manage, govern, and optimize AI grids. While the concept of distributed workloads is straightforward, execution remains challenging- requiring deep back-end integration expertise to bring together the various elements of the AI grid. The Managed Inference Hub has been designed to overcome this challenge. It provides a complete framework for integrating the different elements of the AI grid, curating scenarios for different use cases, and determining which AI models work best for each use case.
The new telecoms paradigm
The traditional telecoms business model is evolving, with operators looking to adapt to new growth opportunities.
AI grids combined with tokenized network capabilities provide an opportunity to accelerate this transformation. Telcos are uniquely positioned to play a leading role in the AI grid landscape. It’s time to consider the art of the possible. Distributed AI inferencing represents the ultimate transition for Telcos from "network as a pipe" to "network as a platform" By focusing on token-based monetization, sovereign edge clouds, and dynamic AI-RAN integration, Telcos can capture high-margin enterprise AI revenues that centralized hyperscaler data centers cannot easily reach.