Navigating the AI hype maze: where will telcos actually find growth?

AI is everywhere, but the real question for CSPs is: where do they fit in?

The scale of the opportunity is significant, where Gartner forecasts worldwide AI spending will reach $2.52 trillion in 2026. Yet, Deloitte’s 2026 enterprise AI research shows that despite accelerating adoption, only a minority of organisations are currently generating revenue growth from AI initiatives.

For CSPs, this is familiar territory. In previous technology cycles, they enabled growth through connectivity while much of the value was captured elsewhere. The AI era creates an opportunity to change that, but only if operators move beyond enabling technology and deliver services customers will pay for.

Why experimentation alone won’t create AI revenue

Across our industry, most AI activity is still in its early stages.

Many organisations – including CSPs - are using AI to improve internal efficiency, automate processes, or enhance existing services. These use cases matter, but they are only the starting point. AI cannot remain an internal capability for CSPs, it needs to be sold.

The next stage is where and how that value is created.

Where CSPs can create value in the AI economy

The opportunity for CSPs is not one-dimensional. There are multiple ways to participate in the AI economy, each with a different level of value and complexity.

At one end of the spectrum, CSPs can continue to focus on connectivity and infrastructure - but as recent experience with 5G has shown, increased demand doesn’t automatically translate into increased revenue. It’s a necessary foundation, but not where most of the AI value will be captured.

Moving further up the stack, operators can begin to monetise AI capabilities by exposing them as services. For example, a CSP could offer fraud detection, network intelligence or AI-powered customer support capabilities through APIs that enterprises integrate into their own applications and workflows.

Beyond that, there is the opportunity to sell AI-driven solutions directly. This is where the model shifts from enabling technology to delivering outcomes. For example, a logistics company may not want to buy an AI agent, but it may pay for predictive fleet optimisation.

At the far end of the spectrum sits the concept of the AI factory - orchestrating an ecosystem of partners to continuously build and monetise new AI use cases. At Beyond Now, we see platform ecosystems and marketplace enablement layers becoming increasingly important - helping CSPs orchestrate partners, accelerate partner onboarding and monetise AI-driven ecosystems at scale.

Each of these positions is viable, but not equally attractive. The challenge is deciding where to focus and committing to the capabilities required to compete.

Why customers buy outcomes, not AI

A consistent mistake in the market is treating AI as the product. Customers buy outcomes.

An enterprise does not want an “agent” - it wants real-time insights, automated reporting or more efficient operations. For example, an AI-driven reporting solution that continuously pulls data from IoT sensors and generates dynamic insights is valuable because it solves a specific problem. This is what the customer pays for.

And this is where the marketplace becomes strategically important. If AI solutions are built from multiple components - models, applications, data, connectivity, cloud, security and partner services - then CSPs need a way to assemble them into something customers can actually discover.

Monetisation at scale requires an ecosystem approach

Turning AI into a revenue driver requires a shift in how services are built, priced and monetised.

AI introduces more dynamic commercial models, from usage-based pricing and API consumption to more granular approaches such as token-based billing or charging per agent run. These models are more complex than traditional telecom pricing, but this is also where CSPs have an advantage. They already understand complex charging, rating, settlement and billing relationships. The task now is making sure those systems are fit for AI-era monetisation.

Value is also rarely created by one player. AI solutions often combine tools and services from multiple providers, making orchestration, settlement and revenue sharing critical.

That shifts the operator’s role from service provider to ecosystem orchestrator - bringing together partners and capabilities to deliver complete, outcome-driven solutions. In that model, marketplaces become more about coordination - enabling AI services, partners and monetisation models to work together at scale.

This is exactly where Beyond Now believes next-generation telco IT platforms will play a critical role - supporting ecosystem orchestration, marketplace enablement and the monetisation of AI-powered ecosystem services.

Business AI in South Africa is an early example of this shift. Launched as Africa’s first vendor-agnostic enterprise AI marketplace, it was designed to help enterprises develop and trial successful AI solutions through a governed marketplace model.

The path forward for CSPs

AI represents a significant shift for telecoms, but it does not come with a predefined model for success.

CSPs have choices to make - where to play, how to build capabilities and how to monetise them. Those that succeed will be the ones that turn AI into clear, outcome-driven services customers will pay for.

Because in the AI economy, the gap between experimentation and execution will define who wins.