“You should drink a shot every time you see a sign promoting telco AI,” joked the CTO of a well-known car company during a presentation at the recent Mobile World Congress in Barcelona. And if people had followed his advice, they would have ended up very drunk. Though primarily a telecoms conference, this year’s Fira […]
“You should drink a shot every time you see a sign promoting telco AI,” joked the CTO of a well-known car company during a presentation at the recent Mobile World Congress in Barcelona.
And if people had followed his advice, they would have ended up very drunk.
Though primarily a telecoms conference, this year’s Fira featured almost as many stands promoting different forms of artificial intelligence, as they did 5G.
While connectivity might still be king in the telco world, AI is the pretender to the throne, threatening to overshadow the monarch.
Much of this pivot has been driven by the demands of telco’s enterprise customers, who want their connectivity providers to offer more than just decent download speeds.
In fact, according to a recent Capgemini survey, early access to cutting-edge technologies and joint efforts on pilots and prototypes is a top expectation that 62% of businesses leaders have of their telco partners.
“In today’s hyperconnected world, telcos are the backbone of the digital economy,” said Praveen Shankar, global telecom leader at Capgemini.
“Businesses expect providers to move beyond connectivity services and offer tailored, end-to-end and flexible solutions that power digitalisation, operational efficiency, and sustainable growth.”
But what role are telco providers playing in getting AI out to their enterprise customers, and how does that fit with the AI revolution going on within their own networks?
Revenue generator
According to PwC, Global telecom revenues are projected to rise at a CAGR of 2.9% through 2028, which is below inflation. The industry will see $200 billion in incremental revenue growth up for grabs, “though 5G revenue remains muted, and cellular IoT services will be buoyed by the automotive sector” the consultancy said in its latest report.
It is no surprise, then, that telcos are turning to AI as a potential revenue generator.
According to Nvidia’s State of AI in Telecommunications report, which surveyed 450 telecoms professionals across the world, over 80% of telcos said that AI is helping increase their company’s annual revenue.
Chris Penrose, Nvidia
“There’s a lot of talk about whether AI investments are truly paying off, but according to the survey, telcos are already seeing significant returns,” Chris Penrose, global head of business development for telco at Nvidia, told TI.
Take French operator Orange, for example, and its enterprise unit Orange Business. Last year, it launched a new multi-LLM solution it calls ‘Live Intelligence’. After road testing the offering with 50,000 internal staff members in France, it took wraps off its so-called “plug and play” service for European customers in November.
To launch the service, Orange partnered LightOn, a French start-up specialising in GenAI solutions while running the offering through the telco’s high bandwidth servers and GPUs, hosted in Orange data centres in France on the trusted Cloud Avenue platform and operated by it.
“Live Intelligence enables all businesses, regardless of their size or sector, to leverage the power of GenAI to improve operational efficiency and customer experience without compromising the security of their data,” explains Orange Business CEO Aliette Mousnier-Lompré. “AI is more than just a technology; it represents a fundamental shift in how we envision future applications.”
At the same time, UK incumbent BT has been developing its AI plans, with BT chief commercial officer Chris Sims underlining the vital role AI will play for its partners going forward.
“The rising tide of AI must lift all businesses,” Sims says. “This technology is no longer the preserve of big corporations with matching budgets. Today’s tools can be set up, implemented, and scaled as needed.”
And in Korea, SK Telecom has reorganised its entire corporate structure around AI, with the aim of making itself into the go-to technology provider for Korean businesses. SKT says its “AI-related revenue grew by 19% year on year” in 2024, though it did not disclose a specific figure.
Traditional telecom services are still important to SK Telecom, but the company’s strategic focus is to “actively pursue the monetisation of AI business” based on the development and growth of AI-centric services, both domestically and overseas, the company said during its recent financial results.
In November, it revealed its AI infrastructure superhighway plan, which focuses on building out a national network of regional AI data centres, enhanced with edge AI infrastructure. It also launched a GPU-as-a-service (GPUaaS) offering for enterprise and government customers in January.
With the GPUaaS now live, it joins AI services SKT is already offering through its four new AI divisions: Adot (A.) SKT’s domestic personal AI agent, which now has 8.3 million users in South Korea, up from 3.2 million at the end of 2023; GPAA (global personal AI agent), which has been developed in partnership with Perplexity, which will go live in the US soon; AIX (AI transformation) is the enterprise and systems integration services unit that will engage with business users in South Korea and overseas; and AIDC (AI datacentre), which includes the already launched GPUaaS offering.
“We strengthened our competitiveness in the telecommunications sector and laid the foundation for our transformation into an AI company,” stated SKT’s CFO, Kim Yang-seob.
“In 2025, we will continue to pioneer the AI era through determination and innovation, further solidifying our corporate value,” he added.
AI as a service
One of the firms working to help telcos provide the right AI services to their enterprise partners is IBM. An IBM Institute for Business Value survey of 300 global telecom leaders found that most communications service providers are assessing and deploying gen AI use cases across multiple business areas.
Sitting down with TechInformed at MWC25, IBM CTO and global industry lead for TME Eoin Coughlan acknowledges that enterprise demand has pushed telecoms firms into positioning themselves as leaders in AI.
“Speaking to industry, they would like an operator to come in and give them not just connectivity, but also the tools to run on top of it,” he says. “They want their telco partners to say ‘here is AI as a service” and to have already vetted a solution that they can go straight to market with.”
One of the key reasons for this is cost. While enterprises could work directly with AI vendors to build a solution, the cost and experience needed to develop a use case from scratch and train the model to solve it can be prohibitive.
“It takes a lot of computing power to train a model from scratch, so what we have done [with our partners] is take an optimise approach. We built a platform, and use open source to help accelerate the uptake of AI.”
This, he adds, allows telecoms partners to tailor the AI offering to the needs of their enterprise base, whilst pairing it with the connectivity solutions they have been building for years.
Nokia VP and global head of sustainability Subho Mukherjee acknowledges that telecoms is at a pivot point, and this is causing suppliers such as the Finnish vendor and its rivals to shift strategy.
“Enterprise is one of our main growth areas,” he says. “In the first 30 years of the mobile revolution we connected subscribers and more digital industries, but now is the opportunity to evolve a lot of so-called ‘physical’ industries.”
In the US, these industries use around 70% of the manpower, supplying jobs across the country in sectors from manufacturing to logistics and energy. “Now is the opportunity to digitise them using communications technologies.”
AI will play a vital role in this because “it has tremendous potential” to transform sectors in a sustainable and manageable way.
AI in the network
For vendors supporting telcos, who in turn are selling direct to enterprise, there are two key elements to AI in communications: How telecoms operators support AI solutions, and how AI can help improve networks.
AI is serving a huge role in the entire lifecycle of the network, explains Nokia Bell Labs president of solutions research Thierry E Klein.
Nokia’s Thierry E Klein
“It isn’t just about designing the network to be more AI native,” he explains. “It is also in the operation of the network. We can bring AI into telecoms networks to optimise their performance, troubleshoot problems automatically, maintain the networks in advance of problems (predictive maintenance) and for customer care purposes.”
He points to two key industrial use cases where Bell Labs and its partners have seen significant growth.
“There are a whole host of areas where we are experimenting with AI in networks, but one of the areas I am most excited about is coupling AI with robots. That requires building a dynamic digital twin of an environment, which captures in real time what is going on.”
So, for example, in factories, robots could become much smarter through AI, able to respond to real, live events, powered by 5G or a private Wi-Fi network. But connectivity is key, and having a smarter network vital.
“Companies, when we talk to them, often say wireless is the problem, but it isn’t,” he adds. “When they didn’t have a reliable network, they just put all of the intelligence in the robot.”
And now that AI can be used to make more reliable networks?
“It’s a challenge to convince them that the wireless network is becoming really mission critical and good enough to take some of the intelligence out of the robot, because now you can optimise your fleet of robots more. I don’t think it’s necessarily a technical challenge. It’s more in the mindset of the industry.”
But while network connectivity is improving, thanks to the advent of technologies such as standalone 5G and 5G private networks, operators still face challenges implementing AI, notably around investment.
The GSMA is the industry body that represents the mobile operator ecosystem and is also the group that runs Mobile World Congress. Alix Jagueneau, who is the GSMA’s head of external affairs, says talk around AI has shifted in the telecoms sphere, from concerns over how it would be regulated, to a focus on innovation.
“We see a lot of countries and governments talking about AI innovation and it is vital for telcos to be a part of that race,” she explains.
“But sometimes maybe they forget that at the very basic level; to compete in AI, we need the mobile infrastructure to be viable and to work. The question of investment becomes very important.”
