- Energy management could be the next big AI use case for telcos
- MTN, AT&T and Deutsche Telekom are looking to AI to cut network power use and improve efficiency beyond what was possible with ML
- Industry execs say the use case is still young, but the upside could be significant
Telco AI adoption started with chatbots, customer service and document summaries as operators sought to tackle the low hanging fruit first. Now, operators are looking to apply AI to a big budget item: energy.
At DTW Ignite 2026 in Copenhagen this week, operators and vendors alike talked up energy management as the next AI use case poised to make a major difference for operators.
“We have demonstrated on a number of sites in South Africa and Cape Town very clearly the world before AI and the world after AI,” MTN Group’s CTIO Charles Molapisi told Fierce at the show. “When we compare the energy savings prior and after, the savings were quite significant because we are now much smarter in terms of what they are processing.”
AT&T’s VP of Network Analytics and Automation Raj Savoor made a similar comment during a recent Fierce Network virtual event, stating that it has been able to leverage network foundation models to deliver a “significant” improvement in energy efficiency compared to its previous ML models.
Deutsche Telekom has also outlined plans to use AI-based algorithms to more efficiently power its network assets on and off.
Still, AI adoption for applications like antenna steering and energy management remains “in its infancy,” AWS VP and telco chief Jan Hofmeyr told Fierce at DTW. But it has the potential to radically change the game.
Today, optimization is done on a time interval basis, he said. With AI, it can be done in or near real time.
“These antennas should be, by default, off, low power, and we power up [when needed]. Today, they are full power most of the time. So, I think there’s a huge opportunity for us to drive efficiency,” Hofmeyr said. “That’s what they need. They need to not deploy more infrastructure. They need to get more out of the infrastructure they’ve deployed before they expand.”
Why it matters and why now
Energy management is a huge opportunity for operators given the amount of money they spend each year on energy. Depending on which source you prefer to use, operators spend an average of anywhere from 4% to 20% of their opex budgets on energy. With both traffic and energy costs rising, it’s entirely possible that that range could grow absent some sort of action.
Thankfully, it seems energy management is one of the easier places to achieve AI-based autonomous operations.
Anders Vestergren, Ericsson’s, head of Solution Area Network Management, BCSS at Ericsson, told Fierce at DTW that when it comes to energy management, it’s occasionally easier to reach Level 4 autonomy “because it’s a single domain” and the technology enabling it is mature and well understood.
Catch all of our coverage from DTW Ignite 2026 right here.