On March 4, the New Communication, New Terminal, New Applicationsummit, hosted by the Global TD-LTE Initiative (GTI) and China Mobile, took place during MWC 2025. At this summit, Chen Haiyong, President of the CS&IMS Domain of Huawei Cloud Core Network Product Line, delivered a keynote speech titled "AI Empowers New Calling, Building the Entry of Intelligent Service." He highlighted that in the 5G-A era, Huawei offers an AI Calling solution that seamlessly integrates with the native dial pad, leveraging audio, video, and data channel capabilities. This innovative solution enables operators to shift from AI-empowered to AI-native, ultimately leading to a more intelligent and seamless communication experience.
Chen Haiyong delivering a keynote speech
Network evolution continues to enhance the user experience. From traditional TDM to all-IP, all-cloud, and now all-intelligent networks, calling services have undergone significant transformations — from analog to digital, voice to video, SD to HD, and now multi-modal communication, redefining the way people interact.
In the age of AI, this technology has become a crucial tool for OTT players, device vendors, and operators seeking to redefine the value chain. New Calling enables seamless task execution through multi-modal perception, intent recognition, and task orchestration, delivering a smarter, more convenient communication experience to a vast user base. This, in turn, enhances user satisfaction and enables operators to establish a key entry to intelligent services.
AI-Powered Calling: Establishing a Service Entry in the AI Era with an Open Architecture
Huawei recommends a two-phase approach for operators to develop an AI service entry.
Phase1: Integrating external AI to enhance calling services. Operators should enable the rapid deployment of large language models (LLMs) with features such as day-level rollout, plug-and-play integration, flexible resource allocation, and multi-model interoperability. By offering seamless connectivity with LLMs, operators can cater to diverse user needs while helping AI models reach a broader audience, fostering mutual growth and success.
Phase 2: Harnessing native AI to transform calling services. Native AI enables seamless fulfillment of complex tasks by integrating multi-modal intent understanding — analyzing voice, visuals, and text — along with intent breakdown, multi-task orchestration, personalized preference memory, number-based call routing, and task self-completion. These advanced capabilities enhance calls with in-call assistance, post-call analysis, on-demand to-do notifications, and more, redefining the communication experience. Furthermore, an open enterprise application ecosystem should be established, allowing third-party developers to integrate and register applications and API services. This will foster the creation of a comprehensive, intelligent communication ecosystem.
AI-Enabled: Rapid Integration and User Attraction with LLMs
New Calling harnesses audio, video, and data channel to achieve multi-modal interactions. By analyzing service scenarios and user subscription profile, New Calling can swiftly interwork with third-party LLMs, fulfilling diverse user needs. For example, New Calling can interconnect with a third-party translation model to provide real-time, accurate multi-language translation services during calls. In everyday scenarios like daily conversations and weather inquiries, it will invoke widely recognized general LLMs to implement quick, intelligent information acquisition and processing. In this way, New Calling not only enhances communication efficiency but also lays a solid foundation for future intelligent communication.
AI-Native: Efficient Completion of Complex Tasks
New Calling streamlines the completion of complex tasks by leveraging multi-modal perception, intent understanding, and task orchestration. This transformation shifts the traditional, inefficiency approach of "one task completed with multiple conversations" to a streamlined method of "multiple tasks completed with a single call." Through number-based interconnection among industrial agents, New Calling helps build an open application ecosystem that allows enterprises to access it without the need for additional adaptation. For instance, in travel planning, the calling agent can help users efficiently complete multiple tasks within a single call, including planning their trips, and booking hotels and flights.
Industry Collaboration Propels the Growth of the New Calling Ecosystem
The rapid advancement of AI will revolutionize the communications industry by transitioning traditional voice and video calls to multi-modal and agent-based communication, thereby enhancing call experiences and efficiency. Supported by AI, New Calling will unlock additional growth opportunities for the communications sector. Achieving this vision requires collaboration among upstream and downstream partners, including terminal and chip vendors, standards organizations, third-party industries, equipment vendors, and operators. Together, we will bridge development gaps, build a new calling ecosystem, and foster the prosperity of the New Calling industry.