Peter Jarich, Head of GSMA Intelligence, presented GSMA’s latest data on AI trends in the telecom industry, covering both current market dynamics and future forecasts for 2025.
Jarich explained that the excitement around generative AI (gen AI) in 2024 has put the spotlight on AI, but this is not new territory for the telecoms industry. He said, “AI is not new for telecoms. We’re all talking about it because of gen AI, but really that’s just a subset of what’s going on in AI.
“We work with operators on a regular basis, and, in our most recent survey, we found around two thirds have an integrated AI strategy. Operators aren’t dipping into AI ad hoc, for individual projects, or at random. They have an integrated strategy that they are now executing.”
What’s more, the majority of operators globally are already providing AI solutions to customers. “Around three quarters of operators are commercially launching gen AI,” said Jarich. “They’re past trials and pilots, and they’re looking at solutions that can support their operations, networks and services.”
Internal and external
One key conversation for the industry is the duality between AI for networks versus networks for AI. Jarich explained, “AI is incredible for improving the way networks operate, for helping operational efficiency, for improving what operators can do with their networks. But, to fully leverage AI, and to roll out the use cases that we want, we need networks that are capable of supporting that.”
This will require capabilities that were not previously required. “If we’re going to have a lot more AI use across networks, there’s going to be a lot more upstream traffic,” said Jarich. “Networks, generally, from a mobile perspective, were built to support more downstream than upstream traffic. So, how do we build networks going forward with those AI workloads in mind?”
Operators do, Jarich explained, understand that they will need to invest in network capacity to support AI usage. And there are currently many questions in the market around how operators are going to make money from AI.
Jarich said, “There’s this internal versus external focus. AI can be used to improve networks, but it can also help with the customer side of things. There’s a tension here, because it’s a lot easier when you’re looking at [AI] internally. You can test out things. It’s a little bit less risky. Operators want to grow that top line, and they recognise you can’t save your way to growth.”
This is where the channel ecosystem could come into play, with operators asking questions around collaboration and reselling. Jarich explained, “Operators are looking to figure out: how do they leverage the innovation taking place across the ecosystem? How do they collaborate with [other stakeholders]? We’re seeing them working more and more with the big cloud players, and this idea of sell through opportunities.
“Operators have a great history of working with the Googles, working with the Microsofts to resell and be a channel for them. We’re starting to see some of this on the AI side as well.”
Ethical and responsible AI
As those market dynamics begin to take root, there is an increasing focus on responsible AI. Alix Jagueneau, head of external affairs at GSMA, and Joaquina Salado, head of AI ethics at Telefónica, discussed the first-hand practicalities and benefits of using AI responsibly.
In September, GSMA published ‘The GSMA Responsible AI Maturity Roadmap’. This was developed in partnership with the GSMA AI for Impact Taskforce and based on insights from McKinsey. It provides companies with a structured framework to establish, monitor and enhance responsible AI practices.
The roadmap is designed to be a tool that allows organisations to assess their current level of responsible AI maturity, identify areas for improvement and align their responsible AI strategies with their ambitions.
Jagueneau explained GSMA created this to be industry agnostic, so the principles can be applied to any sector looking to implement AI in a responsible way. She said, “There is an opportunity to establish trust in the telecoms industry by talking about how responsible AI is not just something that’s good to do. It makes business sense, too.”
She added the roadmap will help operators to put the right governance in place so they can “anticipate potential issues, but also to manage [their] reputation and be ahead of the curve”.
Salado discussed Telefónica’s engagement with the roadmap. She said, “In 2018, we were one of the first companies to establish ethical principles to set the ambition on how we wanted to work with AI.”
She added the company has recently updated those principles because Telefónica wanted to reinforce some key priorities, including sustainability and reducing its environmental impact.
Everyone together
Telefónica, Salado explained, put in place a governance model for responsible AI which includes “roles, responsibilities, and processes” and “reaches out to the whole organisation”. This approach ensures AI is used responsibly and ethically.
By using the GSMA’s roadmap as a blueprint, companies can safeguard themselves against future challenges.
Salado concluded, “There is an opportunity to grow an ethical mindset, because this is about creating cultural change within the organisation. Responsible AI must be integrated completely into the AI strategy. That way, we include the guardrails at the beginning, so we don’t have to rework later.”
This article was included in our January 2025 print issue. You can read the magazine in full here.