AI network capacity demands are set to overwhelm telecom infrastructure, warned Roberto Nobile, CEO of Argentinean operator Personal, during the second keynote at MWC26 Barcelona on March 2, 2026. Nobile said the mobile industry faces a convergence of flat revenues and underinvestment that will soon produce a critical capacity gap.
“Our revenues are flat, and all the investments are far from what is needed today,” Nobile said. “We have a problem, and this problem will get worse. This is what we call the perfect storm.”
AI Network Capacity Could Require a Ninefold Increase
Nobile said artificial intelligence has already created bottlenecks in energy resources during the model training phase. However, he warned that the inference phase presents an even larger challenge for global networks. According to research from Personal’s labs, network capacity will need to increase by nine times to accommodate AI-driven traffic growth.
“It first created a bottleneck on energy resources,” Nobile said. “Models needed all available energy to be trained, but that’s only the training part. Then we have the inference, which is like a snowball. It will also have a huge impact on network capacity.”
OTT Platforms and the Traffic Burden
Nobile renewed calls for over-the-top platforms to contribute financially to telecommunications infrastructure. He noted that more than 75 percent of mobile traffic is concentrated among a small number of hyperscalers. Despite ongoing industry debate, he said operators have made no progress in securing contributions from these platforms.
“We’ve asked the OTTs to fully pay for the increase in data traffic and capacity that they themselves are creating. More than 75 per cent of the mobile traffic is concentrated in a small few hyperscalers. But we have not made any progress.”
Roberto Nobile, CEO, Personal
Autonomous Networks and Architectural Redesign
Nobile argued that AI will embed itself into every application and device, fundamentally reshaping traffic flows and network architecture. He said operators must redesign networks to become autonomous in order to manage the volume of data that AI inference will generate. Nobile described this shift as redefining how networks handle what he called “indirect profit” — AI pushing relevant content to users who then consume it.
“AI is going to be in every app, every machine, everywhere, pushing relevant content to the users,” Nobile said. “This will definitely redefine traffic flows and network architecture, and the designs we need to incorporate AI.”
Consolidation and Spectrum Harmonisation
Beyond infrastructure investment, Nobile called for mandated operator consolidation as a structural response to the capacity crisis. However, he conceded that consolidation alone would not resolve the problem. He also called for collaboration between hyperscalers, terrestrial operators, and satellite operators as a necessary future arrangement.
Furthermore, Nobile highlighted the importance of spectrum harmonisation, greater spectrum availability, and reduced spectrum costs to enable rapid network deployment. The comments reflect a broader industry concern that economic models underpinning mobile networks have not kept pace with the data demands that AI-driven services will impose.
