Generative AI: Reshaping the Telecoms Landscape - Part 2
Part 1 of this 2-part series gave an overview of Gen AI in telecoms and the current use cases of:
1. Customer Services and Chatbots
2. Network Optimization and Predictive Maintenance.
In part 2, I explore three more use cases and give my view of the industry outlook for Gen AI.
1. Personalized marketing and recommendations
Telcos are harnessing the power of Gen AI to revolutionize customer engagement through personalized marketing and recommendations. This shift allows them to move beyond generic campaigns and deliver targeted experiences that resonate with individual subscribers.
This represents a significant shift for telecom operators, as by tailoring experiences to individual needs, telcos can build stronger customer relationships and unlock new revenue streams.
Example: SK Telecom (South Korea)
SK Telecom is undergoing a bold transformation to become a leading AI company. This ambition fuels their innovative AI strategy, exemplified by the launch of ‘A.’ (called ‘A Dot’), a personalized Gen AI assistant service, aiming to revolutionize customer engagement.
SK Telecom uses a multi-pronged ‘AI Pyramid Strategy’ to achieve their AI goals. This strategy focuses on infrastructure development, internal process transformation, and ultimately, delivering AI-powered services to customers. Under this new strategy, the company expects the proportion of AI-related investment to approximately triple from 12% over the past five years (2019-2023) to 33% (2024-2028) over the next five years.
Fig 1: SK Telecom's AI Pyramid Strategy
Source: SK Telecom Gears up for AI Leap in 2024
A. (A Dot) is a personal assistant that sits at the pinnacle of this pyramid, leveraging the company's AI infrastructure and customer data to deliver personalized experiences. Some features that A. (A Dot) delivers are:
- Understanding customer preferences: A. (A Dot) analyzes user data, including app usage, search history, and location information, to build a personalized understanding of individual needs.
- Contextual recommendations: Based on this data, A. (A Dot) can recommend relevant services and products. For example, A. (A Dot) can suggest a new music streaming service based on the user’s recent music app usage or recommend a language learning app based on the user’s travel plans.
- Proactive assistance: A. (A Dot) can go beyond recommendations. It can proactively suggest actions that might improve a user’s experience. For example, A. (A Dot) can recommend an add-on data plan upgrade as the user’s data nears the limit.
Fig 2: SK Telecom is launching A. (A Dot) - a personalized Gen AI chatbot
Source: SK Telecom's A. (A Dot) – a ‘Super App’ Version of ChatGPT
A. (A Dot) is SK Telecom’s answer to OpenAI’s ChatGPT and the telco is looking to integrate its various services, ranging from e-commerce to music streaming, the chatbot. A. (A Dot) is currently only available in Korean. SK Telecom plans to continuously develop A. (A Dot), adding features like English learning and concierge services.
SK Telecom is also exploring partnerships with other companies to expand the service's capabilities and offer even more personalized experiences for customers. During MWC 2024, at the meeting of the Global Telco AI Alliance (GTAA), SK Telecom announced plans to establish an AI-focused joint venture with other partners that include the likes of Deutsche Telekom, e&, Singtel, and SoftBank Corp.
Fig 3: SK Telecom's Global AI Alliance (GTAA)
Through the joint venture company, the five companies plan to develop Large Language Models (LLMs) specifically tailored to the needs of telcos. The LLMs will be designed to help telcos improve their customer interactions via digital assistants and chatbots. The goal is to develop multilingual LLMs optimized for languages including Korean, English, German, Arabic and Japanese, with plans for additional languages to be agreed upon by the founding members.
2. Fraud Detection and Security
Telecom operators face a constant battle against fraudsters who are employing ever-evolving tactics. Gen AI is emerging as a powerful weapon in this fight, enabling proactive fraud detection and safeguarding valuable revenue streams. Especially, in the realm of cybersecurity, Gen AI is instrumental in detecting anomalies and potential security breaches in telecom networks.
Some of the use cases that telcos are enabling in this domain include:
- Identifying anomalies: Gen AI analyzes vast amounts of network data, including call patterns, location information, and billing details. This allows it to identify subtle anomalies that might indicate fraudulent activity, even if it falls outside existing rule sets.
- Predictive analytics: Gen AI can analyze historical data on past fraud attempts and user behavior. This allows it to predict future fraud patterns and proactively flag suspicious activity before it occurs.
- Simulating attacks: Gen AI can be used to generate synthetic data that mimics real-world fraud attempts. This allows telecom operators to test and improve their security systems, ensuring they are prepared for the latest threats.
Example: Airtel (India)
In India, phishing and cyber fraud through unsolicited commercial communication (UCC) have been a major concern for banks, TRAI, and other financial regulators, leading to financial losses estimated at INR 1,000-1,500 crore (~US$120-180 million) every month.
Airtel, one of the largest operators of India, has collaborated with HDFC Bank, one of the largest private banks in the country, to combat phishing scams using the power of Gen AI. This innovative approach tackles a growing concern of SMS-based phishing attacks targeting unsuspecting HDFC customers. This was done by deploying the Airtel IQ Spam Shield, a Gen AI based solution aimed at tackling the rising spam issues.
To achieve this, Airtel's Gen AI model was specifically trained on a dataset of past HDFC phishing attempts. This focused training allowed the model to identify even subtle variations in messaging tactics used by fraudsters.
As a result, Airtel's Gen AI system successfully identified and blocked a large-scale phishing campaign targeting HDFC customers. The messages, disguised as legitimate HDFC communication, contained malicious links that could have compromised user accounts. By proactively blocking these messages, Airtel prevented potential financial losses and protected HDFC customers.
Fig 3: Airtel Gen AI Powered IQ Spam Shield is Protect HDFC Bank Customers from Fraud
Source: Airtel Business
Some of the results that were achieved by this effort were:
- The bank noticed a 98% decrease in spam messages, going from over 2 million spam messages a day to almost none at all.
- The telco managed to block more than 8,000 suspicious SMS headers and prevented over 160,000 fraud attempts by blocking harmful content and templates. Even at 1% conversion, 20,000 consumers being would have been scammed every day if it was not for the Airtel IQ Spam Shield.
- By consistently blocking these messages, the bank is making sure its communication channels stay safe and trustworthy for its users.
The combined efforts of Airtel and HDFC, using Gen AI, created a robust defense against phishing attacks, ensuring customer safety. Proactive detection saved both Airtel and HDFC from the costs associated with responding to and resolving phishing incidents. With continued collaboration, Airtel and HDFC plan to further refine their Gen AI model to stay ahead of evolving phishing tactics.
3. IoT and Smart Connectivity
The rise of the Internet of Things (IoT) and smart connectivity presents exciting opportunities for telcos. However, managing the vast amount of data generated by these devices can be overwhelming. This is where Gen AI steps in, acting as a catalyst for innovation and efficiency, where telecom operators are leveraging Gen AI to manage and optimize connectivity for smart devices.
Some of the benefits of deploying Gen AI for IoT networks include:
- Self-healing networks for IoT: Gen AI enables the creation of self-healing networks for IoT. These networks can analyze anomalies and automatically reroute data traffic if outages or malfunctions occur. This minimizes downtime and ensures uninterrupted data transmission from connected devices.
- IoT device anomaly detection: Gen AI can analyze device behavior, identifying unusual data transmission patterns or sudden drops in signal strength. This early detection of potential device malfunctions allows for preventive maintenance, thereby minimizing disruptions to the entire IoT ecosystem.
- Scalability for future growth: As the number of connected devices expands, Gen AI empowers the network to adapt and scale seamlessly. By proactively identifying and mitigating potential problems, the network can accommodate additional devices without compromising connectivity.
Example: Telefonica (Spain)
AIoT is the combination of Internet of Things and Artificial Intelligence technologies. In this sense, AIoT can be defined as a new technology that provides physical elements with autonomy so that they can analyze the situation and make decisions based on Gen AI.
Telefonica Tech has teamed up with engineering company Grupo Alava to produce a predictive maintenance solution based on IoT, 5G and Gen AI installations.
Predictive maintenance involves using sensors dotted around factories or industrial settings, high speed communications infrastructure, and powerful analytical software to track large amounts of equipment – whether that’s rotary engines, reciprocating machinery, electrical transformers, or even just conveyor belts.
The resulting data pool is analyzed to draw out trends from the production line, and the idea is to spot where there are possible faults or defects in machinery and take measures pre-emptively before the workflow is heavily disrupted.
The platform developed by Telefonica Tech and Grupo Alava also uses Gen AI models designed to improve the efficiency of machinery.
Furthermore, Telefónica has also partnered with Microsoft to integrate ‘Azure AI Studio’ into ‘Telefónica Kernel 2.0’, the company's digital ecosystem, which will allow it to connect its data with Gen AI language models while guaranteeing the privacy of its customers.
Fig 4: Telefonica strengthens its relationship with Microsoft on Gen AI
Source: Chema Alonso, Telefónica Chief Digital Officer
Conclusion and industry outlook for telcos
As we look ahead, Gen AI promises to continue reshaping the telecom industry. Advances in AI algorithms, coupled with the availability of big data and IoT technologies, will further propel the adoption of Gen AI in telecom operations. Future developments may include even more sophisticated Gen AI-driven predictive analytics, autonomous network management, enhanced cybersecurity measures, and hyper-personalized customer experiences.
The convergence of Gen AI with 5G / 5G-A is particularly promising, as it opens up new possibilities for real-time data processing, edge computing (MEC), and immersive services. Telecom operators that successfully integrate Gen AI into their operations will likely gain a competitive edge by offering superior service quality, operational efficiency, and customer satisfaction.
In conclusion, while challenges such as data privacy, ethical AI usage, and regulatory compliance remain, the transformative potential of Gen AI in telecom is undeniable. As innovations continue to unfold, Gen AI will undoubtedly play a pivotal role in shaping the future landscape of telecommunications worldwide.
Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.
Leave a Comment