While AI was as present as ever, a new theme of token monetisation emerged as a key trend of MWC Shanghai. This trend was very minimal at MWC Barcelona earlier in the year and signals the transition of the industry beyond general AI-driven productivity gains toward specific revenue models. Telcos and vendors are now moving beyond monetising bytes to also monetising tokens, a critical currency in the AI era.

Historically, telcos have struggled to monetize new waves of traffic, with video, cloud and mobile apps driving large network demand, but value being captured by internet platforms rather than operators. AI could follow the same pattern if telcos only carry the traffic, rather than owning any of the monetisation opportunities.

Token monetisation is an attempt to create a new commercial unit for AI usage. Instead of charging only for data volume, operators can support AI assistants, enterprise agents and inference services, or guarantee AI service quality. This gives telcos a way to connect network investment with AI consumption, rather than relying only on traditional data plans.

Token monetisation is still early and will not be simple. While tokens are useful for measuring AI workloads, they are quite abstract for most consumers. Mobile and broadband customers understand speed, data allowances and coverage but not token volumes. Consumer-focused AI services will need to be packaged in simpler bundles for specific customer needs. These include AI plans aimed at families, smart homes or productivity tools.

Enterprises and industrial opportunities are much clearer, with customers much more familiar with token-based pricing. For these customers, token-based pricing can be linked directly to output. This is where operators may initially have the clearest path to new revenue. The commercial challenge is still whether operators can provide enough value to justify AI investment. Operators will need to show that AI can increase ARPU, reduce churn, lower operating costs or create new enterprise revenue streams.

Broadband and Wi-Fi become part of AI experience layer

The third theme was the growing importance of broadband and the home network in the AI era. While MWC Shanghai is still a mobile-led event, many of the AI use cases discussed, such as AI assistants, smart home devices, home security and connected devices, will be experienced within the home. Arguably, more complex AI features will be experienced within the home, making the quality of the home network just as important as mobile connectivity. For consumers, the AI experience will be judged by whether the service works reliably, regardless of fiber, FWA or Wi-Fi bottlenecks.

At MWC Shanghai, vendors including Huawei and ZTE positioned their broadband CPE as critical AI enablers within the home. This gives their operator customers control of the home AI experience, with greater visibility to identify faults, manage traffic and improve customer satisfaction. AI features are already being used to manage Wi-Fi, optimise channels, manage mesh access points and resolve issues proactively.

Huawei and ZTE are now enabling their operator customers to deliver far greater value through their broadband CPE. Both are enabling new features such as AI-driven 2D-to-3D video, intelligent motion detection and language user interfaces. Smart hubs, cameras and even pet feeders are being deployed through service providers to grant deeper integration into smart home ecosystems. This increases stickiness, reducing churn and presenting an opportunity to increase ARPU by delivering more value to customers.

As AI becomes more embedded in general smart home products, many will rely on cloud-based models rather than on-device solutions. Even where devices support edge AI, most on display at MWC Shanghai adopted a hybrid approach where heavier AI workloads rely on the cloud. Wi-Fi 7 and future Wi-Fi 8 adoption becomes more valuable in this AI future as the key promise of next-generation Wi-Fi is lower latency, better reliability and improved performance in congested home networks. However, operators still need to balance performance with CPE cost, particularly in a challenging component environment.

Conclusion

MWC Shanghai 2026 showed a telecom industry trying to define its role in the AI economy. While AI dominated the agenda, the key takeaway was how it is being integrated across networks, devices, broadband and emerging monetisation models. Telcos have a clear opportunity to move beyond connectivity, but success will depend on turning AI into practical services that improve customer experience, reduce costs and create new revenue.