MWC24: AI Has Telcos Thinking About the Future of Networks

Streaming Media

AI is set to turbocharge the mobile phone industry making sweeping changes to internal processes and customer applications from telemedicine to personalized live sports broadcasts, according to executives at the Mobile World Congress, Barcelona.

article here

Artificial Intelligence was barely on the radar of the mobile industry a couple years ago. Now, every telco and seemingly every exhibitor at the world’s biggest event for mobile network operators, has AI on the brain.

“AI is set to make sweeping changes to the way telcos conduct their business,” said Mats Granryd, Director General of mobile operators’ body GSMA. “By 2030 AI could contribute $50 trillion to the global economy. As an industry we need to ethically about developing and using the tech for benefit of everyone.”

“AI is the biggest revolution in human history,” said José María Álvarez-Pallete López, chairman and CEO, Telefonica. “For the first time, humans have a technology that can think.”

To underline the point, MWC gave a keynote to Demis Hassabis, the British computer scientist and video game designer who co-founded DeepMind in 2010 and sold it to Google in 2014 for $500 million.

“The advent of AI is an incredible opportunity for mobile and telcos,” Hassabis said. “It will super-charge digital assistants into smart assistants that are actually useful in daily life.”

He questioned, though, whether the mobile phone will be as important a device in a few years as it is today. “AI will make smartphones much more intelligent, but in 5-plus years, is the phone the perfect form factor? Could it be glasses so that AI can see more of the context you are in?”

Álvarez-Pallete López predicted telcos would swiftly move from human-operated networks to one of computer operation supervised by humans. “When you deploy fibre and 5G and switch off the old technology you have something different. The networks become software-based. Then you apply AI and the networks become proactive. It is much more than a telephone network.  This is now a supercomputer we have created,” he said. “The journey to fully autonomous networks is unstoppable.”

“There’s no way we can ignore AI,” agreed Deutsche Telekom CEO Tim Hoettges. “We will see thousands of new use cases in the next few years.” Hoettges said that telcos can maximise labour productivity by up to 37% by embedding AI in their business. AI can help deliver energy efficiency savings, predictive maintenance, and individualised customer service, he said. 

Deutsche Telekom is already deploying AI in 400 use cases across the company and has built an AI “competence centre” of experts developing AI products, he said. It is also a signatory to a pact with 12 other operators including SK Telecom and Singtel to build an LLM for telco specific services.

“We do not want to be dependent on hyperscalers,” Hoettges said. “We want to build our own system without [help from] the outside. I think we understand our world better, so we have to train the LLM ourselves.”

Today, DT makes €10 billion in revenues from European B2B customers. “They desperately want an AI solution for their business,” Hoettges said. “We can be a facilitator.”

Hoettges also emphasised that at DT, AI was a “command strategy from the CEO on down” but that AI skills needed to be upgraded at every level in the organization.

Calling AI a “gamechanger” for this industry, Mike Fries, CEO, Liberty Global said AI was front and centre of his agenda. “Tim mentioned 400 use cases. We are thinking about four,” he said. “It’s impossible to go from zero to sixty and be effective. We aim to be focused,” he explained. Second, while we love to be first, it is more important to be fast. You don’t have to be first in every application or idea. We are feeding this from the top of the company. I am leading it. We are not doing this by the seat of our pants, but in a thoughtful way. We are worried about deployment but also ethics, regulation, insights, and clients.”

One of Liberty Global’s AI focuses is on customer service. Another is the huge area of network optimisation, planning, and energy saving. The biggest is upskilling the company’s staff with tools and mindset.

“Getting AI right is 10% the model, 20% IT, and 70% about people,” Fries said. “The organisational changes required is a massive lift. We want to being everybody along with us.”

Telstra’s ambition is to be the leading “AI-fuelled organisation in Australia,” said Vicki Brady, CEO, Telstra. “Practically, how do we make it happen? Tens of millions of data points cross out networks every five minutes. Humans can’t deal with that. At Telstra we believe it’s got to be a whole business strategy not a tech strategy. Rather than testing things we are embracing AI now. For instance, our customer service teams are using GenAI now get information super quick to customers.”

She said half of Telstra’s processes today are enabled by AI with the goal of 100 percent by July 2025.

While some execs were pondering the vast sums their companies could make by monetizing new AI-driven services, few voiced concerns about its environmental cost. Álvarez-Pallete López said a single AI query “costs the equivalent in power consumption to create one bottle of water or 2 hours of an electric light bulb.” With AI and 5G, he said, “data volumes will rise four-fold between now and 2030 with huge power implications. Massive data from sensors and massive processing capacity presents new challenges including the responsibility of telcos to enhance sustainability.”