AI services offer telcos fresh revenue opportunities – GSMA

Telecommunications operators could unlock new revenue streams by helping businesses and consumers manage, access and optimise artificial intelligence services, moving beyond traditional connectivity offerings, the Global System for Mobile Communications Association has said.

The body said the growing adoption of AI-powered applications, agents and automated workflows was creating an emerging market for telecom operators to monetise AI consumption, with tokens increasingly becoming the unit used to measure and manage interactions with AI models.

In an analysis published on its website, Director of Regional, Social and Policy Research at GSMA Intelligence, Kenechi Okeleke, said operators had largely focused on using AI to improve internal efficiency, but the technology was opening up opportunities to generate revenue directly from customers.

He explained that, just as telecom operators had previously monetised voice minutes and mobile data consumption, the growing use of AI was introducing tokens as another measure of digital consumption.

However, he noted that tokens differed from gigabytes because their value depended on the AI model, the complexity of the task and whether they represented input, output, cached context or reasoning.

According to Okeleke, telecom operators that could simplify the underlying complexities of AI consumption and package services around customer needs would be better positioned to benefit from the emerging market.

The GSMA Intelligence analysis identified three areas where operators were expanding their AI businesses: infrastructure, models and solutions.

At the infrastructure level, telecom operators are deploying data centres, cloud services, fibre networks and edge infrastructure to support AI factories, sovereign AI platforms and graphics processing unit-as-a-service offerings.

At the model level, some operators are developing or partnering with other technology companies to build foundation models suited to local languages and market requirements.

The solutions segment, however, has recorded the strongest operator participation, with more than 200 telecom operators globally already offering enterprise copilots, conversational AI, productivity tools and industry-specific applications.

Asia-Pacific accounts for 43 per cent of global telecom operator participation at the infrastructure level and 38 per cent at the model level, according to the research.

China is emerging as an early test market for the commercialisation of AI consumption by telecom operators. Average daily token usage in the country increased from about 100 billion at the beginning of 2024 to 140 trillion by March 2026, driven partly by the adoption of AI agents and coding assistants.

The country’s three major telecom operators are experimenting with different commercial models to capture the growing demand.

China Telecom bundles access to AI models with connectivity and cybersecurity services for consumers and businesses, while China Mobile combines AI tokens with cloud PCs and AI agent capabilities in regional packages.

China Unicom is also exploring a credit-based model that allows customers to manage token consumption across multiple AI models.

• 2027: IG, INEC commissioner to discuss AI misinformation at PUNCH webinar
• Meta deploys AI to detect underage users
• Varsity to secure NUC’s AI approval

Okeleke, however, cautioned that China’s experience might not be directly transferable to other markets, given the country’s substantial cloud businesses, investments in AI infrastructure and broad ecosystem of open-weight models.

He said telecom operators elsewhere would need to develop commercial strategies that reflected their infrastructure, customer base and competitive environments.

Beyond reselling access to AI tokens, the analysis identified the control layer as a potentially more defensible source of revenue for telecom operators.

This involves helping businesses manage AI services across different models and providers, including selecting suitable models for specific tasks, routing workloads, controlling costs and meeting governance requirements.

The report identified model aggregation, intelligent routing and workflow optimisation as three stages through which operators could develop these capabilities.

Model aggregation involves bringing different AI models together through a single interface, application programming interface and billing system. Intelligent routing allows customers to select models based on cost, quality, speed, availability and regulatory requirements, while workflow optimisation focuses on translating AI adoption into measurable productivity improvements and cost savings.

Citing industry examples, Okeleke said China Mobile’s MoMA platform could reduce token costs by between 20 and 30 per cent through intelligent routing, caching and context reuse.

He also referenced AT&T’s reported savings of up to 90 per cent from its internal AI gateway, arguing that telecom operators could demonstrate the commercial value of such platforms within their own operations before offering them to customers.

The analysis said operators had existing advantages that could support their entry into the AI control market, including established customer relationships, billing and identity management systems, security and compliance expertise, local regulatory knowledge and distributed infrastructure.

However, it warned that telecom operators faced growing competition from hyperscalers, cloud service providers and AI-native companies already developing platforms to control access to AI services.

Okeleke said operators needed to identify their positions in the AI value chain, develop relevant capabilities and establish partnerships without becoming low-margin resellers of AI services.

He also urged them to prepare their networks for the changing demands of AI traffic, including potentially higher uplink requirements and more distributed processing, using network application programming interfaces and edge capabilities to differentiate their offerings.

The executive warned that the opportunity for telecom operators to capture a meaningful share of the AI economy was narrowing as technology companies moved quickly to establish their positions in the market.

According to him, operators that failed to act risked remaining connectivity providers while companies controlling the intelligence and service layers captured a larger share of the economic value generated by AI.