Network and service assurance Value: Reduce cost to serve, differentiate customer experience AI and genAI solutions enable networks to be smarter, more reliable and environmentally sustainable while reducing operational burdens. Autonomous networks powered by AI and genAI can transform every step of network management, spanning design, deployment, engineering, orchestration and operations across public, private 5G and programmable networks. Design automation optimizes network designs to maximize coverage while minimizing capital expenditure (CapEx), supporting smarter decision-making. GenAI enhances this process by automating documentation and speeding up planning approvals. Digital twin planning creates virtual network models, enabling data-driven decision-making and efficient resource allocation. AI tools like deployment quality assurance (QA) assistants drive efficiency in deployment. The assistant provides real-time quality assessments to enhance quality control, streamline review processes, reduce costs and deliver an improved customer experience. In network engineering and orchestration, AI- powered automation enables dynamic, real-time management of multi-vendor, multi-domain networks. By automating integration, reconciliation and standardization of network data, AI accelerates vendor onboarding, improves compatibility and minimizes manual interventions. GenAI solutions like the network engineering assistant use a chatbot interface to help engineers monitor network infrastructure by analysing real-time network performance data to provide insights and recommendations that drive network optimization, performance and user experience. Once live, intelligent network operations shift from reactive to proactive management through AI-driven operations centres that predict potential faults before they escalate. By autonomously initiating resolutions, these centres minimize downtime and enhance service reliability. An AI-powered network operations centre (NOC) genAI assistant automates key tasks such as incident troubleshooting and change management for NOC engineers, improving efficiency and reducing operational costs. This use case accelerates network operations, enabling reliable, scalable, cost- effective management and advancing towards fully autonomous network operations. In network decommissioning, genAI simulates scenarios to predict network impacts, optimize CapEx and redistribute resources optimally. It automates technical data reconciliation, updates technical records and generates scripts to streamline the shutdown of outdated equipment and setup of new infrastructure. Throughout these processes, a genAI-powered field tech assistant can provide technical assistants on the field with real-time installation and troubleshooting advice. AI also drives sustainability by reducing network energy costs, 87% of which is consumed in the radio access network (RAN)22 in mobile networks. The RAN energy savings assistant analyses network data to identify energy-saving opportunities. Combined with 3D digital twins, these insights enable network engineers to simulate and implement energy-saving strategies that reduce energy costs by up to 15%23 without compromising performance. Identifying “sleeping cells” BOX 4 Rakuten Mobile tackled the industry-wide issue of “sleeping cells” (malfunctioning cells which appear operational on management systems) by deploying an AI-powered detection algorithm. This system analyses network key performance indicator (KPI) data in near real time, using a hybrid dynamic global auto-encoder to identify anomalies. The initiative reduced detection latency by over 80% and improved productivity by more than 60%. By shifting from reactive troubleshooting to proactive detection, Rakuten Mobile achieved greater network reliability and cost savings, underscoring the critical role AI plays in the evolution of modern telecom networks. Telefonica’s AI-powered digital twin BOX 5 Telefónica Germany’s AI-powered digital twin solution enhances transport and IT infrastructure management. It creates a dynamic virtual replica of the physical transport network, allowing precise simulation, prediction and optimization of operations. By integrating advanced AI and ML models, it analyses real-time and historical data to identify performance bottlenecks, forecast demand and streamline maintenance. This innovation has reduced troubleshooting times by up to 90%, improved reliability and positioned Telefónica as a leader in sustainable telecom practices. GenAI simulates scenarios to predict network impacts, optimize CapEx and redistribute resources optimally. Artificial Intelligence in Telecommunications 14

Artificial Intelligence in Telecommunications 2025