Tokens touted to telcos as ticket to monetising AI – starting in China

Details: John Tanner; Category: Enterprise Ecosystems; 25 June 2026; 333 views

A hot topic at MWC Shanghai 2026 is AI tokens that enable telcos to monetise AI services. Chinese telcos are already doing it – but that market's unique environment raises questions over how telcos outside of China can follow their lead.

At a summit session on Wednesday, GSMA CTO Alex Sinclair said that as telcos evolve into the so-called “IQ era” – in which AI will reshape the mobile sector from connectivity provider to intelligence value provider – the combination of 5G-Advanced and 6G connectivity, reliable telco-grade AI and a collaborative ecosystem that unlocks true scale enables telcos to monetise AI with token-based services.

“These token-based approaches are an important first step not just in pricing AI services, but in redefining how value is measured, exchanged, and scaled across the ecosystem, and they create a real opportunity for operators to move beyond connectivity to become providers of integrated AI solutions,” Sinclair said. “Operators are not just enabling AI, they're becoming a core part of the AI economy.”

For the uninitiated, an AI token is the smallest unit of data processed by AI models during training and inference. This gives service providers a measurable way to determine pricing of AI services based on the value of the application as well as the resources needed to deliver it.

This is key because the resources involved aren’t just network capacity – they also include compute power, electricity consumption, model and platform, all of which will vary from application to application. Put simply, AI tokenisation enables telcos to calculate the intelligence value of an AI service based on those various metrics, and bill for them accordingly, said Cao Lei, deputy MD of China Telecom’s 5G Innovation Center.

“For example, the computing power is a very important part of determining the cost, and the model sets the ceiling for how far the AI can go,” Cao said. “The platform determines the efficiency, and the application determines the value.”

Token-based AI plans in China

Because this means the pricing will vary based on usage of resources, AI tokenisation might sound to some like a new and more complex version of metered billing for data – a concept telcos have struggled to monetise before – but it may also be the easiest way for customers to understand it.

That’s the approach China Telecom, China Mobile and China Unicom have taken as the first telcos in the world to launch token-based AI plans, which package access to AI usage in a similar way to traditional mobile data bundles, said Pablo Iacopino, head of research at GSMA Intelligence.

“Chinese operators are targeting B2C as well as B2B for AI tokens with obviously different pricing models,” Iacopino said during the same session. “What is really important is that AI tokens create a commercial link between AI demands and AI infrastructure investment.”

This matters because many operators are already investing in AI technologies (or planning to do so), at least for their own internal use – AI tokens give them the opportunity to transform that investment into new revenue streams.

“We see more than 65% of operators have an AI strategy, and are going to implement that strategy,” said Wang Yongde, president of Huawei’s Carrier XtoB business. “AI is going to drive the new growth from B2B, from traffic to token.”

Eric Yang, president of Huawei’s Carrier business unit, added that the vendor is seeing plenty of demand on the enterprise side, especially from SMEs.

“Artificial intelligence is becoming a universal need for companies, so this is becoming a very important engine of growth for operators,” Yang said.

Huawei – whose booth at this year’s MWC Shanghai event is dedicated largely to promoting technology solutions from its portfolio that can support AI tokenisation – expects that by 2028, AI will be in high demand in enterprise sectors including retail, education, healthcare and manufacturing, among others.

“We know that our connectivity, reliability, security, and [partner] cooperation, as well as our computing power and algorithms, are very, very good. So we can provide one-stop services for all of them,” Yang said.

Will it play outside of China?

The question, of course, is whether operators outside of China can replicate what China Telecom, China Mobile and China Unicom are doing with AI tokenisation.

GSMA Intelligence’s Iacopino is optimistic. “China is leading on this, but we believe that other operators in other countries will follow this path, because it's part of a broader push by operators to commercialize artificial intelligence, cloud, and compute capabilities,” he said.

Iacopino added that demand will also be driven by the fact that telcos inside and outside of China are already reporting revenue generated from AI. For example, China Mobile has reported that AI is already generating 9% of total revenue, while for South Korea's SK Telecom, that figure is 4%.

“That means that the AI opportunity is real in some markets, and operators are really, really keen to drive monetization,” Iacopino said.

That said, it also helps that Chinese telcos are adopting AI as part of the 15th Five year Plan implemented by the Chinese government. Earlier this month, Bloomberg reported that the government is preparing to launch a US$295 billion AI infrastructure plan to build data centres across the country, with Chinese telcos operating them and Huawei providing most of the AI chips.

Meanwhile, in a separate keynote session on enterprise AI at MWC Shanghai on Thursday, Oscar Ramos, MD of Orbit Ventures, observed that one reason AI tokenisation could gain traction in China’s B2B market is that many enterprises there have typically not taken to subscription models for performance-based services.

“They don’t mind to pay more as long as they gain more value out of that,” he said. “So you need to figure out a way to attach the way you charge to the value you create to the savings that you generate.”

It also has to be done transparently, especially for SMEs in emerging markets, due to the complexity of the tokenisation model, Ramos added.

“The token economy is very complicated, and because so many SMEs are the backbone of the economy in most emerging markets, you cannot go to many of these people and start talking about token economy,” he said. “So pricing these services in a way that makes sense is critical.”