Telco Giants Embrace AI to Power New Revenue Streams – MWC 2024

Executives from some of the biggest players in telecom believe AI could provide opportunities for new revenue streams and also redefine how they interact with customers and partners.

In a panel discussion at Mobile World Congress 2024, Lilac Ilan, global head of business development for the telco market at Nvidia said telcos could generate revenue by monetizing idle capacity for AI inferencing.

Generative AI systems need consistent compute to generate outputs. Ilan’s idea could see telcos offer inferencing capabilities as a service to businesses and developers whose systems need low-latency processing but lack IT infrastructure. This could enable them to run their AI models at the edge of the network.

She offered statistics from a recent Nvidia report, where 19% of respondents who said they were adopting AI reported revenue increases of almost 10%. Moreover, she said that monetizing capacity at the edge concept would result in “much bigger” increases.

One method for monetizing networks some telcos are doing at present are to offer sovereign cloud services to enterprises, she said.

Use case impact

Telcos are using AI today. Nokia CTO Jitin Bhandari said his firm already has “real use cases” that utilize traditional AI models.

These include applying AI to improving customer experiences and running optimizations as well as use cases related to energy savings and cybersecurity.

Related:Huawei Builds an AI Model for Telcos – MWC 2024

However, the advent of generative AI represents a shift for telcos. Bhandari explained that Nokia ran an open source large language model on nodes of its MX Industrial Edge solution to make it more conversational and “easier to connect to these industries and enterprises.”

AT&T uses AI to better understand outages in hurricane scenarios, said Adam Loddeke, assistant vice president of RAN technology at AT&T. He said AI can provide performance and operational benefits for telcos.

But the changes brought about by AI do not occur overnight. Jean-Christophe Laneri, head of cognitive network solutions at Ericsson, said that telcos need to evolve their network “step by step” to enable these capabilities.

Ericsson believes AI will change automation: “We are at a bit of a critical point now, where you have the complexity on one side, the cost pressure, the complexity of these networks. And at the same time, you have the readiness and maturity of the technology, like the cloud and AI,” Laneri said.

“I think we are at a critical point here where we can actually do something really interesting.”

Partnerships to power AI

Related:How Telcos are Using Generative AI - MWC 2024

A consensus among the panel was that coming together could empower telcos to further benefit from AI.

A day before the panel, a new initiative was launched called the AI-RAN Alliance, where major telcos and cloud providers pledged to enhance mobile network efficiency with AI.

Ericsson was among those who joined the project.

“It is going to take a village,” Laneri said. “If you look at AI transformation in telcos, … It's a lot about process changes, rationalization, ways of working and so forth. This is going to take an entire industry to pull up, it's not a one-vendor show.”

AT&T was not among those who signed onto the project, but Loddeke said that AI has brought America’s largest wireless carrier together with partners whom it previously never would have interacted.

“You can't walk through [Mobile World Congress] without pretty much every place you walk in, somebody's got something on AI. And so, I do think it creates a lot of opportunities,” he said.

“Once you open up the network and you have the capability to onboard new applications that are there built by anybody in the room and you can take advantage of best of breed, you have the advantage of putting that into your network and running that and so it's going to drive a lot of innovation competition.”

Skill issues

Bhandari from Nokia said that the ecosystem needs to come together to help understand the technology.

He said that people and skills need to be encircled by a governance framework to drive that understanding through the formation of things like model strategies.

Nvidia’s Ilan referred back to Nvidia’s survey, which found that 55% of telcos said that skills were a bottleneck to scaling AI. “We have a great opportunity in front of us to further skills in our teams and employees to just adopt it faster.”

Both Loddeke from AT&T and Ericsson’s Laneri said building trust was a vital part of building out employee skills.

“The skill sets may exist within our own teams, but you have to transition that knowledge downstream within your organization so they are comfortable with the new technology that's moving forward,” Loddeke said.

“We need to create that trust,” Laneri added. “We're not going to get into a fully autonomous system without having the trust of the engineering teams. The algorithm’s outputs need to be explained, not the black box, so that we get that trust, and step by step, automate the workflows.”