i卡會員

In recent years, the telecom and mobile communications industries have made significant investments in the 5G spectrum and its infrastructure and have been pushing for its commercialization. However, overall returns have yet to meet expectations. Under the dual pressures of stagnant pricing and required ongoing infrastructure investment, the topline growth of the global telecom industry remains weak. The total revenue in 2024 decreased by 1% from 2023, and the average revenue per user (ARPU) has continued to decline.

To revitalize the industry, Chairman of the GSMA, Mats Granryd, proposed the "GROW" strategy and suggested that telecom operators should focus on the deployment of 5G Standalone (SA) networks, proactively explore new use cases to boost revenue, leverage innovative technologies such as Open Gateway and AI to develop new markets, and adopt forward-looking planning for future spectrum resources to ensure ongoing innovation and growth. According to GSMA, the Open Gateway initiative launched in 2023 has attracted the participation of around 80% of telecom operators around the world, indicating a high level of willingness for collaboration. The Open API market is projected to reach US$300 billion and could expand into verticals such as automotive, aviation, and finance. The integration with AI will further accelerate API commercialization. 

Against this backdrop, AI has become the driving force for innovation of communications technology. The integration of AI and mobile communications is deepening, not only in product innovation but also for business model transformation. Huawei has launched the industry's first AI-powered core network; Samsung introduced AI camera features for its Galaxy S25; and MediaTek showcased its AI for the high-speed interconnectivity of data centers. Telecom operators are also keen to deploy generative AI in customer service, network management, and internal operations, and to expand and diversify revenue streams with AI assistants, e-commerce platforms, and enterprise applications.

While AI has yet to become fully mainstream, open network architecture remains a closely watched direction of development. After years of technological evolution and practical validation, open architectures are now an important option for 5G deployment, with a foothold in the private network market and its ecosystem continuing to expand. Japanese electronics giant Kyocera announced its entry into the 5G base station market and teamed up with companies from Taiwan, South Korea and India to form the O-RU Alliance at the MWC 2025 to strengthen cross-border industry collaboration for open networks. 

Meanwhile, AI technology has demonstrated huge potential in network energy saving and performance enhancement. With AI/ML working at the physical layer, 5G transmission power can be reduced by up to 50%, significantly improving energy efficiency yet maintaining the transmission quality. At the MAC (media access control) layer, AI can automatically adjust signaling and access strategies based on business needs, enhancing network flexibility and agility. 

Not only industry heavyweights but also startups are rushing to integrate AI and mobile communications. At the MWC 2024, the AI RAN Alliance was launched by multinationals including NVIDIA, AWS, Arm, SoftBank, Ericsson, Nokia, Samsung, Microsoft, and T-Mobile, igniting a wave of enthusiasm for deploying AI in radio access networks.

Telecom operators also see AI as a key driver for operational transformation and the enhancement of service monetization capabilities. In 2024, Deutsche Telekom, Emirates Telecommunications, Singtel, SK Telecom, and SoftBank signed a joint venture agreement to co-develop a multilingual large language model (LLM) for telecommunications. Its successful deployment in intelligent customer service systems showcases the benefits of AI implementation.

AI has also become central to the business models of startups. According to MWC’s startup platform 4YFN (4 Years From Now), about 30.4% of the startups participating at the 2025 exhibition see AI as key to their value proposition, and another 29.3% are using AI to optimize products and services. This highlights the innovative potential and industrial momentum driven by the integration of AI into technologies in various fields.  

In the integration of AI with RAN (radio access network) and other infrastructure, telecom operators have achieved concrete results in three aspects, all closely related to the enhancement of customer experience. First, in terms of network deployment, AI can analyze 5G load and environmental conditions to precisely select the best location of base station sites, optimize Massive MIMO Beam Steering, improve signal coverage, and predict the operation mode of base stations based on user demand to further save energy and boost efficiency.

Second, AI-driven network slicing technology allows telecom operators to create multiple logical networks on a shared infrastructure architecture, catering to the needs of different industries and use cases and delivering flexible and customized services. Third, in terms of operations and customer services, AI significantly improves efficiency and user satisfaction. 

While AI is all the rage across the industry, telecom operators must also carefully manage potential risks. Currently, most AI use cases are about internal efficiency optimization and cost control, without clear, scalable business models yet. If unable to stay on top of AI platforms, telecom operators risk being reduced to mere "transmission channels" again, with profits captured by AI and cloud platform providers — or worse, facing marginalization within the industry.