Telcos will rely heavily on artificial intelligence and AI agents to meet user demands, increase operational efficiencies, carry more traffic, and improve services.
Telcos are on the cusp of an evolution, with once-unimaginable trends rapidly becoming commonplace. It comes as no surprise that taking centre stage among trends is artificial intelligence (AI) and more specifically AI agents (Large Language Models (LLMs)), as well as AI-driven tech services. In this article, Harsha will explore six pervasive telco trends, including AI chip penetration, the Intent-Net, the continuing rise of data and device usage, the driving force behind new business models, satellite mobile telephony, and combating fraud with AI machines.
1) The rise of AI chip penetration
In 2023, AI chips were in around 4-5% of smartphones; today, it’s approximately 16%, and by 2028, it’s predicted to be approximately 50%. When equipped with an AI chip, mobile devices can build knowledge around preferences and requirements, remember interactions, and accomplish specific tasks. As LLMs become increasingly agile, they will be much easier to run on mobile devices. However, making this a reality requires telcos to change the way they approach network design, both in terms of how the mobile interaction happens and how the customer sees the result.
A massive amount of knowledge resides in the mobile, which gives telcos the opportunity to provide hyper-personalisation. Using LLMs to scan the data, service providers can uncover unique patterns about the user. While telcos have historically been slow in adopting recommendation engines, they are now beginning to view their networks as an AI entity – and with good reason. They already have the network elements, the customer data, and the churn data; the next step is for telcos to train an AI model on top of the information to better understand how the network is performing and what can be recommended to deliver better customer experiences.
While the penetration of AI chips has begun, there remains some resistance to the use of LLMs in the cloud due to privacy issues. However, with the emergence of open-source AI such as Llama, LLMs can be hosted at the edge of the network, decreasing privacy concerns.
See also: Telcos Turn to AI to Solve Their Biggest Problems
2) Intent-driven Network enablement
This trend requires two parts: the Intent and a Network. The Intent-Net provides the means for users to interact with the ecosystem – both on the customer experience side and on the network side. From a telco perspective, they gain a better understanding of the customer’s Intent. Using GenAI recommendations, which will flow into the network to be orchestrated to meet individual customer expectations, service providers gain the ability to provide customers with the hyper-personalisation they increasingly seek.
This process begins with AI agents. They will ‘see’ the Intent and determine how to facilitate the request. Acting autonomously, AI agents determine what the Intent of the user is and the steps required to successfully accomplish the task. Intent-driven Networks provide telcos with the ability to deliver results much faster than traditional human-driven software ever could.
In practice, an Intent-Net consumer use case example may be something like: “We want to go to Australia next month – find us flights, accommodation, restaurants, and sightseeing excursions.” The Intent is the interaction the consumer has with their mobile phone. The rest falls upon the AI agents and apps on the network to accomplish the task. A more complex use case could be prioritising network resources for a drone data transfer. This example includes network slicing, opening up APIs, integrating LLMs, and so on. Delivering on these types of use cases requires a variety of technology elements, which many of today’s more sophisticated telcos already possess.
See also: GenAI – The Telco Differentiator Poised to Drive Personalized User Experiences
3) The continuing rise of data and device usage
Although many European and American markets are relatively saturated in terms of subscriber growth, content usage continues to grow exponentially. Additionally, there remains developing markets where people do not yet have a mobile phone. As new subscribers come on board, data growth and device usage will continue to exponentially increase.
In today’s data-hungry world, it’s all about short videos, TikTok, Instagram, and YouTube, and telcos are continuing to witness huge growth in the use of data. For many consumers, the time spent on smartphones typically means watching videos. Video watching already tops five hours a day, and this figure is only increasing.
See also: How Telcos Are Overcoming GenAI Challenges to Realize Its True Value
4) The driving force behind new business models
New business models are largely being driven by the enterprise sector. With a wealth of potential to be exploited, this trend is delivering unprecedented value for telcos in the following areas:
- As 5G SA networks become even more pervasive, they will have a significant impact on the enterprise sector. For example, T-Mobile’s network slicing for first responders delivers priority service, lower latency, and faster speeds.
- With the arrival of 5G RedCap, the emergence of new IoT use cases is set to increase. Although restrained by reduced capability, 5G RedCap provides telcos with an extraordinary amount of network flexibility, enabling them to deliver less complex yet differentiated services.
- Across geographies, 5G Fixed Wireless Access(FWA) continues to deliver a remarkable amount of success.
- The focus on fiber, combined with merger and acquisition activities surrounding fiber, continues to take priority. Users want a seamless experience, and telcos realise that the bulk of internet traffic is through these WiFi broadband networks, particularly the heavy workload video applications. Telcos are seeking ways to capitalise on that traffic and move away from simply being a wireless provider.
- Fintech is emerging as a business in its own right. Originating as a small team within the telco, Fintech are increasingly becoming a full business in emerging markets, and telcos are increasingly procuring independent fintech applications. Global players like Orange, which is European-based, have expanded their fintech presence in developing markets, reflecting a growing trend of diversification within telecom, especially in developing regions.
5) Satellite mobile telephony
Originally against the tide, telcos are increasingly adopting satellite mobile telephony. While questions around licensing remain, Airtel and Jio have partnered with some of the low Earth orbit satellite companies and are experiencing success in this area.
6) Combating fraud with AI machines
Increasingly, telcos are leveraging AI machines to combat a variety of fraud types, such as device fraud, GenAI fraud, subscription fraud, and account takeover and identity fraud.
Across regions, device fraud is the leading use case that telcos are dealing with today. Consumers pilfering mobile devices continues to be a big and profit-draining issue. Deepfake audio and video are increasingly getting democratised, making GenAI fraud increasingly more prevalent and more newsworthy. Subscriptions have become the norm for many, increasing the amount of money flowing through subscription-based channels. There’s a lot of interest and deployment of anti-fraud measures in this space, and with enterprise rollout, strength in cybersecurity is becoming even more critical. Using techniques such as SIM swapping and cloning or exploiting devices for illegal gain, account takeover, and various identity frauds are now considerable use cases across geographies.
Making inroads against fraud, AI agents are now equipped with tools, knowledge, and memory to battle a variety of fraudulent activities. The impact LLMs have against fraud reaches from mobile phone automation through to BSS/ OSS automation. Using AI agents, telcos now have previously unattainable opportunities for fraud investigation. This will put them at the forefront of reshaping how the industry thinks of software and architectures.
