13 November 2024

Classical and GenAI are reshaping telecommunications as we know it – so what do operators need to know?

AI is significantly reshaping the telecommunications landscape, from network optimisation and customer experience through to security and new business models.

“AI is already transforming the telecommunications industry, enabling telcos to enhance service quality, boost customer satisfaction, and rapidly introduce innovative products and services,” shares Dominic Smith, Marketing Director, Cerillion. “Many telecom companies are now leveraging AI-driven chatbots and virtual assistants to deliver immediate customer support, efficiently handling routine issues through automated digital channels, while allowing customer services personnel to focus on more complex or high-value tasks.”

“The technology shows potential to significantly improve network management and optimisation. Many of our customers have already told us how AI-powered solutions are empowering them to enhance their network performance, increase operational efficiency, and elevate customer experiences,” adds Lucky La Riccia, Vice President and Head of Cloud Software and Services at Ericsson Middle East and Africa.

Reshaping the landscape

With AI, telcos can automate routine tasks such as traffic routing, fault detection, and network maintenance, while AI-enabled predictive maintenance can anticipate hardware failures and optimise performance, reducing downtime.

“We have seen firsthand how AI can be an ally to MNOs in tackling their most persistent challenges. By deploying more intelligent solutions, MNOs can optimise network resources dynamically in real-time, effectively managing congestion and ensuring seamless connectivity,” says La Riccia. “AI-driven analytics can proactively identify customer churn risks. This allows operators to engage at-risk customers before they leave. Meanwhile, machine learning models can analyse transaction patterns in mobile financial services to detect anomalies in the fight against fraud, thereby significantly reducing potential losses.”

“Telcos consistently face the issue of stranded and underutilised assets based on the network footprint,” adds Seshan Krishnamurti, Vice President - Sales, Africa Region, Covalensedigital. “AI can help with asset lifecycle management from the viewpoint of recovery, and redeployment (e.g. upgraded enterprise CPEs) as well as faulty asset valuation (though it is not monetisation, it can help avoid write-downs and any associated market impact).”

The rollout of 5G, too, is deeply intertwined with AI, which can be used to manage the complex and dynamic nature of 5G networks, including massive data flows, multiple connected devices, and diverse application requirements. And, as more data is processed at the network edge, AI is used to make real-time decisions in applications like autonomous vehicles, smart cities, and industrial IoT, which require ultra-low latency and high reliability.

“For 5G to make sense, digital transformation use cases from various industry sectors are essential and AI is at the heart of this transformation, touching every business priority across customers, products, operations, employees, ESG,” explains Krishnamurti. “In telecoms, AI has delved into the world of network operations and customer engagement where various data elements related to behaviour are used to train algorithms - initially for detection and prediction and in some cases trigger relevant action.”

“As more countries transition to 5G - with its faster speeds and reduced latency - telecom operators are being presented with new complexities. Today’s AI technologies can transform these complexities into opportunities, bringing together big data with unique network domain expertise to deliver unprecedented benefits for network operations and more,” says La Riccia.

Making money

One of the most hotly anticipated advantages of AI adoption for MNOs is monetisatoin. AI can analyse individual customer data to offer highly personalised services, such as custom data plans or premium content packages, leading to higher customer engagement and increased average revenue per user (ARPU).
Indeed, Krishnamurti believes that AI can help MNOs “understand customer behaviour nuances to tailor product/service offerings and the experience. Creating a higher bundle with immersive experience in gaming with the associated higher pricing, device sales, etc., personalises the offering to the affluent; while the same technology can be used for immersive learning offered to the less affluent through community centres from a wholesale perspective. The first option offers higher ARPU while the second one creates higher volumes, offsetting the lower margins.”

“With smartphone usage surging, operators can utilise AI to analyse network traffic and application data to create personalised offerings that drive customer engagement and support higher retention rates,” agrees La Riccia. “AI also enables dynamic pricing models based on real-time demand and network resource availability. This helps operators to maximise their revenue potential. Moreover, the integration of AI can facilitate the development of more innovative services for both consumers and enterprise customers. By embedding AI into traditional service offerings and operations, operators can ultimately achieve new business models while lowering operational costs.”

According to Smith, many telcos struggle to turn a marketing proposal for a new product or promotion into the necessary Business Support System (BSS) configuration required for monetisation: “in fact, numerous ideas never progress beyond the planning stage because the process of building, testing and launching new products is slowed down by outdated systems and workflows. GenAI is a game-changer, enabling telcos to develop and launch products and services much faster by bridging the gap between marketing and operations. It allows companies to move directly from brainstorming to product testing and validation, using natural language and image recognition to create catalogue configuration within seconds. With GenAI, CSPs are now fully equipped to compete with the digital service providers by embracing a ‘fail fast’ approach to launching and monetising new products and services.”

Implementation obstacles

While AI has the potential to significantly improve business strategy and operations for MNOs, factors like technical and financial barriers and organisational and cultural resistance are hindering adoption.

“Operators need to evaluate whether their current infrastructure is ready to support AI functionalities, which may involve investing in computing and storage capabilities,” confirms La Riccia. “Cybersecurity and regulatory compliance are equally critical as operators must safeguard consumer data amid an expanding digital landscape. Operators would also do well to establish trust in AI through explainability and human oversight. Incorporating mechanisms that promote trustworthiness will enhance stakeholders’ confidence in AI adoption, especially as the user base grows.”

Data remains a key barrier to AI adoption, too. AI relies on high-quality, large datasets to train models and generate insights. Some MNOs may face challenges in collecting, curating, and managing the necessary data for AI solutions - and inaccurate, incomplete, or siloed data can result in poor model performance.
Indeed, “the speed of change in the digital era requires strategy to be fluid and adaptable; the single biggest asset allowing for this flexibility is data and the quality of that data,” shares Krishnamurti. “In many areas of the world, the quality of the data suffers from being inaccurate, average-poor quality and not always easily accessible.”

“One challenge is the gap between some companies’ digital transformation ambitions and their capacity to realise those ambitions quickly. This gap can be driven by macroeconomic challenges that delay investments, or the lack of ecosystem collaborators necessary to spur innovation. Evolving regulatory environments and the lack of adequate ICT infrastructure - especially in rural areas – can further exacerbate such issues,” confirms La Riccia.

AI implementation and maintenance also require specialised expertise in data science, machine learning, and AI technologies. Alas, there remains a global shortage of such skilled professionals, and MNOs may struggle to hire or retain talent capable of managing complex AI projects.

“Moreover, competition for the skills comes not only from within the industry but across multiple industries; in the short term, skills movement, knowledge transitions create gaps, not easily filled,” agrees Krishnamurti.

To support the widespread adoption of AI within the telecommunications ecosystem, there is a need to invest in training and reskilling the current workforce, which requires time, resources, and a willingness to adapt to new ways of working.