Cerillion believes telecom operators could play a growing role in the commercial delivery of artificial intelligence services, creating new requirements for the billing, charging and partner-management systems that sit behind them.
The immediate focus is AI tokens. These are the units used by large language models to process and generate information, and they provide a measurable way to track AI consumption. That has led to comparisons with traditional telecom units such as minutes, messages and mobile data.
Some operators are already testing the model. China Mobile, China Telecom and China Unicom have introduced AI token plans covering multiple AI providers. This suggests telecom companies are looking beyond basic connectivity and considering how they can package, manage and charge for AI services.
Cerillion argues that selling token bundles directly may only be an early stage in that development.
Customers are unlikely to care how many tokens an AI service consumes. They are more likely to pay for a useful outcome, such as an AI assistant that handles emails, organises travel, performs research or supports customer service. The token usage would remain in the background. That could make the underlying billing infrastructure more important than the token itself.
Telecom operators could buy or manage AI capacity from different providers, package that capacity into customer-facing services and use their BSS/OSS platforms to track usage, calculate charges and settle payments. Customers would buy the service, while the operator manages the complexity behind it.
Tokens could become cheaper and less visible over time. Telecom history provides a clear precedent. Individual voice minutes, messages and units of data became less important to customers as networks improved and bundled services expanded.
AI tokens could develop in the same way. Different models also consume different numbers of tokens for the same task, which makes a simple token allowance harder to understand than a conventional data plan. Cerillion therefore sees a broader opportunity in enabling AI service monetisation.
If operators offer AI services from multiple providers, they will need systems capable of tracking usage across different models and services. Those systems may need to handle token volumes, model selection, latency, service levels and other charging variables before calculating customer bills and payments to partners. That creates a more complex commercial environment than traditional telecom billing.
An AI service may cost more depending on the model used, the amount of processing required or the speed at which a response is delivered. Autonomous AI agents could add another layer of complexity if they perform multiple tasks across different platforms and providers. These services would require flexible product catalogues, real-time charging and accurate settlement between operators and AI suppliers.
Cerillion already provides software designed to manage charging, billing, customer relationships and partner arrangements. If telecom operators become intermediaries between AI providers and end customers, those capabilities could become increasingly important.
Cerillion plc (LON:CER) is a leading provider of billing, charging and customer management systems with more than 20 years’ experience delivering its solutions across a broad range of industries including the telecommunications, finance, utilities and transportation sectors.
