Deutsche Telekom exec: AI poses massive challenges for telecom industry

Deutsche Telekom’s VP of technology strategy, Ahmed Hafez, co-hosted the DSP Leaders World Forum 2023 session entitled “Creating a framework for the AI-native telco” this week in the UK.  He said that AI will deliver the telecom sector its biggest ever challenges and opportunities, but to take advantage of the benefits that AI will bring the industry needs to figure out a way to evolve from being opportunistic to becoming AI-native.

To date, the telecom sector has been exploring the potential of AI without looking at the bigger picture, and that holistic view needs to be taken in order to figure out the best way to go, Hafez believes.

Like so many other pundits and cheerleaders, Hafez regards the impact of AI as “the biggest transformation we will ever encounter.” And this is not only about the magnitude of what AI will do, but also the pace – it will outpace our understanding of things so fast, so we need to be ready…

“Previous transformations have [happened at an] accommodating pace – they were not changing so fast that we couldn’t comprehend or adapt to them. In order for us to adapt to AI, we need to transform as individuals, not [just as] companies. On an individual level you need to be able to comprehend what’s going on and pick the right information.”

To illustrate the magnitude of the challenges that AI will deliver to the telecom sector, Hafez presented a few supporting statistics:

  • The AI market was worth $136bn in 2022 and is set to be worth $1.8tn by 2030
  • The telecom AI market alone was worth $2.2bn in 2022
  • Global private investment in AI reached $91.9bn in 2022
  • AI delivers a 40% increase in business productivity, according to a study by Accenture (Hafez thinks that number is too low, that productivity gains will be much higher)
  • There are already thousands of AI-focused companies – by 2018 there were already nearly 3,500
  • AI will drive the need for 500x compute power between now and 2030 (“What does that mean for telcos? How can we deal with that?” asked Hafez)
  • In terms of human resources, 63% of executives believe their biggest skills shortage is in AI expertise
  • Three in every four CEOs believe they don’t have enough transparency when it comes to AI and are concerned about skewed bias in the AI sector

So a lot of eye-opening trends that should give the telecom industry food for thought, especially when it comes to attracting employees with AI skills. “How will we get the people we need if there are thousands of AI companies” attracting the experts, he asked.

Hafez also related how he encountered what he described as some “depressing” information about how unattractive telecom operators are to potential employees, especially those of a younger generation. Of the top-50 most attractive companies in advanced economies for employees, none of them are telcos: “This is a worrying trend… we need to become more attractive to the younger generations,” he noted.

The telecom industry began exploring the use of AI in earnest less than 10 years ago, noted the DT executive, when it started looking into its potential with proofs of concept and trials. “Then we took the opportunistic approach to AI – use case-based, where you find a good use case, you implement it and it’s concrete. There’s nothing bad about that, as it’s the right thing to do… and we’ve been doing that for a while and it’s delivering value. That’s fine as long as you are doing a few tens of use cases.”

But using AI at scale, which is what the industry needs to do to become AI-native, where AI is fully integrated into everything and becomes part of all operations and decision-making processes, throws up a lot of new questions about how the sector progresses from being opportunistic to becoming AI-native – what are the missing steps, Hafez asked?

Source: Deutsche Telekom

“Once we start to ask, what would the future be with AI in everything we do, in every appliance, in every application, in every network component, it would be over the top. You would have data that is being worked on by five or six AI engines, creating different things…. You would have not just tens of use cases, but hundreds, or thousands. Are we prepared for that? Are we ready to embrace such scale? Are we building AI for scale? I don’t think so.

“We are building AI trying to get things done – which is okay. But in order for us to get through this journey, through this transformation, what stages do we need to pass through? What are the steps that we need to take to… make sure that the problem is clear. If we have a huge amount of AI, do we run the risk of conflicting AI? So if I have AI for energy efficiency and I have another one that actually improves network quality, could they create conflicts? Can they be a problem? If I have AI that is on the optical layer and AI on the IP layer, can they make different decisions because they consume data differently?

“If we look at things from this perspective, do we need, within our organisations, another stream of hiring people and the need to upskill leadership? Do we need to upskill ourselves to help our teams? What do we need to do? If you look at technologies, do we need to change the perspective of how, for example, the 3GPP is building the standards in order to make sure the standards are AI friendly? Do we need separate standard bodies to look at AI? What would be their functions? What would be their scope?” asked Hafez.

And does the industry need a framework that can provide guidance so that the telecom sector can develop in the same direction with its use of AI?

“This is the discussion we want to have, and I hope the message is clear – we have a great opportunity, but opportunities do not come without challenges,” he cautioned.

Hafez set the scene for a great discussion with his fellow speakers, Juniper’s chief network strategist Neil McRae, Rakuten Symphony CMO Geoff Hollingworth, Nokia’s CTO for Europe Azfar Aslam, and Digital Catapult’s CTO Joe Butler – and it’s fair to say there were differences of opinion! You can view the full session on demand here.


Here are some specific examples of how AI is being used in the telecom industry in 2023:

Network optimization:

AI is being used to analyze data from network sensors to identify potential problems before they occur. This allows telecom providers to take proactive steps to fix problems and prevent outages. For example, companies are using AI to predict network congestion and proactively reroute traffic to avoid outages. 5G networks began to roll out in 2019 and are predicted to have more than 1.7 billion subscribers worldwide – 20% of global connections — by 2025.  AI is essential for helping CSPs build self-optimizing networks (SONs) to support this growth. These allow operators to automatically optimize network quality based on traffic information by region and time zone. AI in the telecom industry uses advanced algorithms to look for patterns within the data, enabling telecoms to both detect and predict network anomalies. As a result of using AI in telecom, CSPs can proactively fix problems before customers are negatively impacted.

Customer service automation and Virtual Assistants:

AI-powered chatbots can answer customer questions and resolve issues without the need for human intervention. This can free up customer service representatives to focus on more complex issues. For example, Verizon is using AI to power its Virtual Assistant, which can answer customer questions about billing, service plans, and technical support.

Predictive Maintenance:

AI-driven predictive analytics are helping telecoms provide better services by utilizing data, sophisticated algorithms, and machine learning techniques to predict future results based on historical data. This means operators can use data-driven insights to monitor the state of equipment and anticipate failure based on patterns. Implementing AI in telecoms also allows CSPs to proactively fix problems with communications hardware, such as cell towers, power lines, data center servers, and even set-top boxes in customers’ homes. In the short term, network automation and intelligence will enable better root cause analysis and prediction of issues. Long term, these technologies will underpin more strategic goals, such as creating new customer experiences and dealing efficiently with emerging business needs.

Robotic Process Automation (RPA) for Telecoms:

CSPs have vast numbers of customers engaged in millions of daily transactions, each susceptible to human error. Robotic Process Automation (RPA) is a form of business process automation technology based on AI. RPA can bring greater efficiency to telecom functions by allowing telcos to more easily manage their back-office operations and large volumes of repetitive and rules-based actions. RPA frees up CSP staff for higher value-add work by streamlining the execution of complex, labor-intensive, and time-consuming processes, such as billing, data entry, workforce management, and order fulfillment. According to Statista, the RPA market is forecast to grow to 13 billion USD by 2030, with RPA achieving almost universal adoption within the next five years. Telecom, media, and tech companies expect cognitive computing to “substantially transform” their companies within the next few years.

Fraud Prevention:

Telecoms are harnessing AI’s powerful analytical capabilities to combat instances of fraud. AI and machine learning algorithms can detect anomalies in real-time, effectively reducing telecom-related fraudulent activities, such as unauthorized network access and fake profiles. The system can automatically block access to the fraudster as soon as suspicious activity is detected, minimizing the damage. With industry estimates indicating that 90% of operators are targeted by scammers on a daily basis – amounting to billions in losses every year –  this AI application is especially timely for CSPs.

Revenue Growth:

AI in telecommunications has a powerful ability to unify and make sense out of a wide range of data, such as devices, networks, mobile applications, geolocation data, detailed customer profiles, service usage, and billing data. Using AI-driven data analysis, telecoms can increase their rate of subscriber growth and average revenue per user (ARPU) through smart upselling and cross-selling of their services. By anticipating customer needs using real-time context, telecoms can make the right offer at the right time over the right channel.



6 Common Uses of AI in Telecommunications

Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases

Allied Market Research: Global AI in telecom market forecast to reach $38.8 by 2031 with CAGR of 41.4% (from 2022 to 2031)

SK Telecom inspects cell towers for safety using drones and AI

The case for and against AI in telecommunications; record quarter for AI venture funding and M&A deals

Global AI in Telecommunication Market at CAGR ~ 40% through 2026 – 2027

Cybersecurity threats in telecoms require protection of network infrastructure and availability

China to launch world’s first 5G cruise ship via China Telecom Corp Ltd Shanghai Branch


Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>