MTN’s AI RAN Bet: Turning Cell Towers Into Smarter Infrastructure Hubs
MTN Group is positioning itself for a future where mobile networks don’t just carry data, but actively learn and optimise in real time. A key part of that shift is MTN’s investment in a new AI RAN specialist called ORAN Development Company (ODC), a move the company says fits squarely inside its Ambition 2030 strategy.
In an interview with TechCentral, Mazen Mroué, CEO for digital infrastructure at MTN Group, explained that the idea isn’t simply to adopt AI tools as add-ons. Instead, MTN wants AI to be built into the radio access network itself—an approach the operator believes can reduce costs and better match conditions across Africa.
The investment involves a group of technology and network players including Nvidia, Nokia, Cisco, AT&T, Booz Allen, and Telecom Italia, alongside MTN. While the headline is the funding round, the deeper story is MTN’s goal to evolve from a traditional network provider into a broader platform business—where technology innovation and network capabilities create new value.
AI in the RAN: MTN’s push to cut costs and improve efficiency
A radio access network, or RAN, is the part of a mobile network that connects devices to the operator’s infrastructure using radio spectrum. It sits closer to the user experience than core network functions, handling the “last mile” of connectivity between phones and base stations.
MTN’s argument is that this layer is where AI can deliver fast, practical gains. By embedding intelligence directly into RAN operations, the company believes it can reduce the need for expensive data centre capacity for certain workloads run by external parties.
Mroué pointed to a familiar challenge for operators and service providers alike: the tendency to build more capacity in response to demand, even when not every workload truly needs that level of compute.
“We cannot continue building data centres without looking at ways to optimise capacity and efficiency,” Mroué said. His view is that many tasks still require connectivity and transport, plus other processing steps across the network.
By targeting AI closer to the edge—meaning at the RAN level—MTN expects that some responsibilities involving transport and core processing could be avoided for specific use cases. The result, he suggested, is cost-effective connectivity and a more efficient way to support applications without waste.
MTN’s broader Ambition 2030 plan has three pillars: fintech, digital infrastructure, and the connected home. On the digital infrastructure side, MTN is also investing in terrestrial fibre and plans to expand into AI data centres in both South Africa and Nigeria. So the strategy is not “AI everywhere” in a generic sense; it’s more like choosing where AI belongs to produce measurable outcomes.
ODC’s technology is also tied to MTN’s longer-term network evolution pathway. The company is preparing for 5G, but Mroué framed the investment as part of anticipating the bandwidth and complexity that may come with 6G and future high-demand applications.
Why MTN invested in ODC instead of buying technology
ODC’s funding round totals $45 million, but MTN did not reveal how much it personally contributed. Mroué said MTN could have chosen the simpler option—buying AI RAN technology “off the shelf.” But the operator argues that direct investment provides something more valuable: influence over how AI RAN is shaped and adapted.
That matters because network conditions in Africa are not identical to those in other regions where much of the underlying tech is developed. Mroué emphasised that MTN wants its voice included in shaping the technology so it better fits real operational environments.
“Not everything that’s developed in the West can really suit and meet the needs of our continent,” he said, adding that MTN values the wide range of contributors involved in the initiative. The group includes chipmakers, mobile network equipment manufacturers, and operators—covering the ecosystem needed to make the solution practical.
He also noted that MTN’s focus is not only about serving third parties. AI in the RAN, he said, can help optimise the network itself as workloads fluctuate and as system complexity increases over time.
That idea aligns with broader industry concerns. Network management is becoming harder because every generation of mobile technology adds layers of configuration and decision-making. In a white paper titled “Autonomous intelligence in RAN,” Ericsson argues that the number of parameters operators need to manage is growing rapidly, making traditional rules-based approaches less effective.
The paper highlights a step-change pattern: 2G networks involve hundreds of tuneable parameters per base station, 3G doubles that level, 4G adds more, and 5G ramps it significantly higher—while 6G is expected to require even more. With that growth, manually tuning networks becomes increasingly unrealistic.
MTN’s push comes as competitors explore similar paths. TechCentral previously reported that Vodacom partnered with Nvidia to build an AI-powered digital twin of its RAN in Cape Town, designed to help manage complexity across multiple connectivity generations and device types.
Vodacom’s group technology strategy lead, Ryan van den Bergh, described how quickly networks become complicated when they must handle everything from older 2G feature phones to advanced 5G smartphones, using different components and behaviours across the network.
MTN’s ‘unique environment’: powering towers efficiently with AI
Mroué said MTN’s challenges include more than just software complexity. The operator also operates within a “unique environment” where grid power is not guaranteed, meaning towers often require efficient self-powering to maintain service.
In that setting, AI-enabled optimisation isn’t just a performance upgrade—it can be a practical necessity. If workloads and network demands can be managed more intelligently at the edge, MTN’s infrastructure can potentially use power more efficiently while still meeting connectivity expectations.
By investing in ODC, MTN is effectively betting that the next major evolution in mobile networks will come from intelligence being embedded into network architecture—not only added in the form of separate platforms and data-centre-based processing.
For now, MTN isn’t publishing detailed performance targets or timelines for ODC’s deployment. But its message is clear: AI RAN is central to lowering costs, improving efficiency, and ensuring that mobile networks can evolve in ways that fit Africa’s operational realities.
MTN’s AI RAN push signals a shift from “running networks” to building adaptable network platforms—and with data centres under pressure and network environments varying widely across the continent, the company’s bet on edge intelligence could become a defining strategy in how African operators modernise next.