Yue Wang, chief technologist at China Telecom, said that AI-native 6G will require a fundamental architectural shift toward the joint orchestration of networking and computing resources
In sum – what to know:
AI-native vision – China Telecom said future 6G networks must jointly orchestrate communications, compute, storage and AI resources to support emerging AI-driven services.
Cloud convergence – The operator said integrating network and cloud infrastructure, along with organizational changes, is essential to enable autonomous AI operations.
Service evolution – China Telecom expects future AI applications, including physical AI and robotics, to drive demand for low-latency, compute-aware 6G capabilities.
Current telecom networks were not designed to support the requirements of future AI services, according to Yue Wang, chief technologist at China Telecom, who argued that AI-native 6G will require a fundamental architectural shift toward the joint orchestration of networking and computing resources.
Speaking during RCRTech’s Telco AI Forum, Wang said today’s telecom systems remain largely based on deterministic architectures built around predefined interfaces, rules-based logic, and human-engineered control mechanisms.
According to Wang, AI-native networks should be viewed across three layers: infrastructure, operations, and services. “The first is the infrastructure layer. So this is where the network computes storage AI resources needs to be planned and deployed. And then the second layer is the operational layer. This is where you optimize and orchestrate all the resources. And then the third is the service layer. This is where the network can support the AI services,” she said.
Wang also said that future AI applications will require much more than connectivity. “In the future, AI services will not only ask for connectivity, but they will also ask for compute latency adaptation at the same time,” the executive said.
While operators are increasingly deploying AI capabilities into existing systems, Wang argued that this incremental approach has limitations because current networks were not originally designed around AI workflows.
“The data is not AI ready. The APIs are not always designed for closed loop control. Network functions are not designed with, the AI lifecycle management, real-time decision making, etc.,” the China Telecom executive added.
Rather than simply adding more AI software, Wang said the industry must rethink how network and computing resources are managed together. “So I think it’s not a lack of AI hours you put into a software system, but rather it’s the ability to orchestrate the network and compute together to support the new AI services we face today,” she said.
Wang also stressed that the transition toward AI-native networks will require changes from both the telecom and AI communities. “AI must adapt to telecom requirements, because network is a critical infrastructure. We can’t allow uncontrollable decisions, for example, in a live carrier network,” she said.
At the same time, telecom architectures must become more flexible to accommodate AI-driven operations. “We should make AI reliable enough for telecom. But at the same time, we need to allow telecom architecture to be flexible enough to benefit from AI,” Wang added.
Wang also said it expects emerging AI applications — particularly physical AI systems such as robotics and industrial automation — to become important drivers for AI-native 6G capabilities. “I think the first capability, if we are thinking of 6G networks, is native to the AI services. I would say it’s the communication and the computer orchestration jointly,” she said.
Looking three to five years ahead, Wang said success in AI-native 6G will depend on the industry’s ability to create a flexible platform capable of exposing and orchestrating multiple capabilities simultaneously.
“We will know that we are on the right path of AI native 6G when the network can seamlessly expose and orchestrate the communication, compute, data, intelligence, and assurance as the capabilities to our customers,” she said.
