Telecoms operators, no stranger to change brought about by rapid technology transformation, are facing unprecedented profitability pressures from rising inflation, diminishing revenues and increasing personnel, energy, capital and operational expenditures. At the same time, telecoms operators typically produce and have access to large datasets and possess an opportunity to automate what are often systemic manual processes at scale. Many telcos are turning to AI solutions to reduce costs, create value and drive efficiencies. The global market for AI in telecoms is tipped to reach $39b by 2031, driven by smartphones with AI onboard, generative AI powered by large language models, AI-driven network management and hyperscale computer and storage availability on demand.

AI-powered chatbots are being used to increase customer satisfaction by reducing wait times, providing context-driven real-time responses and improving experience personalisation. In 2022, chatbots were estimated to have saved industry approximately $11b in part by enabling enhanced customer self-service and reduced agent hand-off and truck rolls. AI chatbots not only understand and respond to customer queries but also feed into operator’s documentation and record-keeping processes. For example, Vodafone has deployed its chatbot TOBi, powered by Microsoft Azure AI, which handles 25-30 million conversations each month in 16 markets and in 15 languages with the expectation to scale to 500 million conversations. Rakuten has partnered with OpenAI to change how consumers shop and engage with merchants using conversational AI experiences.

Nokia predicts that networks will grow by 73% in the next five years, which is five times more than the rate of growth over the previous five years. Networks are also growing in complexity as well as size, given the deployment of new and evolving network technologies, which also raises the likelihood of network performance issues and increased energy costs. To combat these increasing costs and complexities, Ericsson forecasts that AI-powered network solutions can result in a 35% reduction in critical incidents, 60% reduction in network performance issues, and 15% reduction in energy costs. Hutchinson 3 Indonesia deployed an AI-driven network performance solution from Nokia to optimise RF coverage by identifying gaps and interference, delivering 17% higher spectral efficiency and 60% faster network optimisation through automation compared to legacy tools. BT recently unveiled plans to cut up to 55,000 roles, from the current 130,000 by the end of 2030, with at least 10,000 of those job cuts due using technology such as AI-driven automation to realise network and operational efficiencies.

Fraud, unauthorised network access and reactive maintenance cost telecoms operators billions every year. The Communications Fraud Control Association has reported that telecommunications fraud amounted to $40b in 2021 equivalent to 2.22% of total industry revenue. Telecoms operators are now leveraging AI and machine learning to replace in some cases largely manual processes instead with large dataset analysis to identify patterns, detect anomalies and empower decision machines to prevent adverse outcomes from occurring. For example, Nokia launched its AVA Fixed Network Insights solution using Bell Labs developed AI/ML models that analyses data collected regarding access and WiFi networks, routers, device and OSS to provide automated decision making to proactively identify and resolve broadband problems, reducing cost and increasing customer satisfaction.

A number of operators are collaborating with industry to provide network infrastructure to specifically support AI use cases and workloads. DriveNets introduced its AI network solution DriveNets Network Cloud-AI designed to maximise the utilisation of AI infrastructure and improve the performance of large-scale AI workloads. Nvidia and SoftBank Corp announced their collaboration on a pioneering platform for generative AI and 5G/6G applications, which involves building data centres that can host generative AI and wireless applications on an Open RAN server platform.

AI solutions deployed by telecoms operators must ensure the operator’s compliance with any existing obligations in any authorisation or licensing conditions regarding the provision of electronic communications networks and services, as well as other areas such as consumer protection, security obligations, competition and open access obligations. This will require in depth understanding any AI/ML decision mechanism to ensure compliance, such as having the relevant controls in place for a chatbot providing a responses or guidance to customer queries in a manner that involves different management and governance considerations from human-involving processes. Adopting AI solutions will also introduce further obligations on operators to comply with AI specific regulations and standards, as well as additional data privacy, security, and ethical considerations.

The AI powered revolution presents telecoms operators and their service providers with significant challenges and very clear opportunities to improve efficiency, performance and profitability while at the same time raising novel and unchartered regulatory and policy issues. We continuously monitor regulatory developments affecting telecoms operators and service providers  over the world, including those regarding deploying and supplying AI driven products and services. Get in touch if you’d like to have a further discussion about your AI related projects and we’d be delighted to assist.