AI is already improving telecom operations. But turning those capabilities into services customers will pay for requires a different approach.

“We already have an AI platform.”

It’s something we hear regularly in conversations with telecom operators. And they’re right. AI is already helping operators optimize networks, automate customer care, predict service disruptions and reduce operating costs.

These investments are important. But improving operational efficiency is not the same as building a new revenue stream.

For operators looking beyond connectivity for growth, the question is no longer whether AI has value. It’s how to turn that value into something customers are willing to buy.

AI Adoption Is Growing. Revenue Has Yet to Catch Up.

According to McKinsey’s 2026 research, AI contributes less than 2% of total revenue for more than half of surveyed operators that have tested AI use cases. Yet 64% of telecom leaders expect AI to contribute more than 5% of revenue in the near future.

The figures reflect AI’s broader financial contribution, not just revenue from customer-facing AI services. Still, they highlight the difference between what operators are achieving today and what they expect AI to deliver. 

Omdia’s July 2025 research projected telco B2B AI revenue of $4 billion for 2025, with a forecast compound annual growth rate of 65% through 2030. This includes AI-related connectivity, hosted infrastructure, platforms and professional services.

The forecasts are encouraging, but revenue will not follow automatically from AI adoption. Operators still need to decide which services customers will pay for, how to take them to market and who will own the commercial relationship.

We believe operators have an opportunity to participate in this growth by making AI-enabled services available directly to their business customers.

Operators Don’t Need to Build Every AI Service They Sell

Telecom operators have spent decades building the commercial infrastructure needed to deliver services at scale.

They already manage enterprise relationships, sales channels, recurring subscriptions and billing systems. These capabilities provide a foundation for introducing services beyond connectivity.

But having the infrastructure to sell a service is not the same as having a service that businesses want to buy.

Operators need to identify practical customer needs, select relevant AI applications and build offerings that are straightforward to understand, purchase and use.

Developing individual AI applications from scratch is not necessarily the most practical route. It requires specialized expertise, ongoing investment and the ability to keep pace with rapidly evolving technology.

An alternative is to bring together relevant AI applications, package them around customer requirements and offer them through an operator-managed services platform.

This allows operators to concentrate on the areas where they already have experience: managing customers, delivering services and building recurring commercial relationships. Technology partners can provide the underlying AI capabilities, while operators retain ownership of the customer proposition.

The distinction matters. Operators do not have to become AI model developers to build an AI services business.

What Would Operator-Led AI Services Look Like?

Consider a small accounting firm handling hundreds of invoices and financial documents every month. Its employees spend considerable time extracting information, checking entries and transferring data between systems.

AI-powered document processing could reduce that manual effort. But adopting it independently means identifying a suitable provider, understanding the pricing, managing another subscription and potentially arranging technical integration.

For a business without dedicated IT resources, those steps can become barriers to adoption.

An operator-led AI services marketplace could offer a simpler route.

The firm could discover a document processing service through its existing telecom provider, choose a subscription based on its requirements and receive the charges on its regular telecom bill.

If it later needs an AI customer service assistant or workflow automation tool, it could add those services through the same marketplace rather than starting a separate procurement process.

The model is not limited to voice assistants. We see opportunities across several practical business functions: 

The operator’s value is not necessarily in developing these applications. It is in making them easier for businesses to access, subscribe to and manage.

For operators, the opportunity is to build recurring subscription revenue by extending an established customer relationship into a new service category.

Selling AI Services Requires More Than a Marketplace

A catalogue of AI applications is a starting point, not a complete business model. The harder work is making those services relevant to customers and giving them reasons to subscribe and continue using them.

That begins with the problem being solved. A small business may not be looking for an AI platform, but it may be willing to pay for a service that reduces the time spent processing invoices or handling customer enquiries. The commercial proposition needs to make that value clear.

Pricing and packaging matter just as much. Customers need to understand what is included in a subscription, how charges are calculated and what happens when their requirements change. Familiar billing arrangements can simplify the purchase, but they cannot compensate for a service that is difficult to understand or use.

Operators also need the flexibility to introduce new applications as the market develops. A service portfolio that works today may need to change as AI capabilities improve and customer expectations evolve.

And there is an organizational consideration. Building an AI services business requires more than technical readiness. Product, enterprise sales, billing and customer support teams all have a role in making the model work.

These are commercial challenges that telecom operators understand. Applying that experience to AI services could be as important as the technology itself.

Agentrix: Building the Commercial Foundation for AI-as-a-Service

At Telcovas, we’re developing Agentrix around this opportunity.

Agentrix is our AI-as-a-Service platform, designed to help telecom operators offer AI-enabled applications to their enterprise and business customers through an operator-managed subscription model.

Our focus is on the commercial infrastructure needed to make that possible: bringing AI services together in one marketplace, enabling customers to discover and subscribe to applications, and connecting subscription charges to the operator’s existing billing ecosystem.

Rather than asking operators to build and maintain every AI application, we’re developing a model that allows them to bring relevant services to market through their established customer channels.

We are developing Agentrix to support applications across customer service, document intelligence, workflow automation and business analytics, with the flexibility to expand the portfolio as customer requirements evolve.

Agentrix is still at an early stage. Our immediate focus is on building the capabilities that can help operators establish and grow a customer-facing AI services business.

Explore Agentrix, Our AI-as-a-Service Platform →

The Next Step Is Turning AI into a Business

AI will continue to play an important role in making telecom networks and operations more efficient. That investment remains essential.

But there is a separate opportunity emerging outside the operator’s own operations: helping businesses use AI without requiring them to navigate a fragmented market of technologies, providers and subscriptions.

Operators already have much of the commercial infrastructure needed to serve that market. What they need is a practical way to bring together the right applications, make them easy to buy and manage them as ongoing services.

We believe that is where AI monetization can become a meaningful part of the telecom growth strategy.

The industry has made considerable progress in using AI to run its business. The next challenge is learning how to sell AI as a business.