Growing AI Data Traffic Requires More Licensed Mid-Band Spectrum

The rapid expansion of artificial intelligence (AI) will drive a surge in data traffic across mobile networks. A recent Accenture report highlighted the risk that limited 5G network capacity due to insufficient spectrum availability means over a third of data traffic from AI-enabled devices will go unmet by 2029. More licensed commercial spectrum is needed to cope with anticipated AI data traffic.

AI is poised to contribute up to $15.7 trillion to the global economy annually by 2030-more than the current output of China and India combined. But securing a competitive AI industry transcends economic opportunity-as the Wilson Center put it, "AI is not just a technological race-it is a contest of governance models that will shape global power, economic growth, and individual freedoms." Policymakers are rightly focused on supporting U.S. AI leadership, in turn aiming to extend our economic edge and technological superiority in a way that supports American values.

The future of mobile AI depends on spectrum.Accenture, The Looming Spectrum Crisis

Unfortunately, Accenture's study shows the AI opportunity is at risk because of a lack of full-power mid-band spectrum to support the anticipated AI-traffic boom. Accenture isn't alone with this concern. The CEO of DigitalBridge expects mobile data traffic to grow three to five times over because of generative AI. Nokia is "preparing for a surge in 5G uplink traffic." Ericsson analyzes possible AI futures, all of which see steep growth in mobile traffic even despite countervailing efficiency gains from AI used within the network (for example through better video compression technology). The CEO of Vodafone has warned that cellular networks could be "the next AI bottleneck."

5G mobile networks are key to AI's continued evolution, allowing users to bring AI resources to bear on the world away from their desktop. While early AI tools are largely text-based, future iterations will offer audio and video interactivity. As Accenture explains, "AI tools will be able to take in surroundings captured by wearables or the user's device, and process queries in real-time, while producing video or audio responses back tailored to the user's situation." This transition to AI-based video processing in the cloud will see mobile traffic explode.

Powerful AI tools integrated into various consumer applications will make 5G use-cases like augmented reality far more compelling. Take, for example, a plant identification application-various tools can analyze a photo of, say, a curious flower you see on a walk, and tell you what kind of plant you're looking at. Today it's a bit clunky and doesn't always work. But imagine with high-speed 5G and AI-based image processing running in the cloud-hold your phone up to your field of view and every plant species you see is instantly identified.

Similar tools combining AI with 5G will scale up smart factories and industrial internet of things. Technicians will be able to find and fix problems with complex production systems using augmented reality, navigation applications will better adjust to real-time traffic updates, distributed logistics and supply chain management will be more efficient, and AI will help optimize wide-area smart city services. The promise for productivity gains is tremendous, but experts worry our mobile networks aren't ready for the onslaught of additional data traffic.

Accenture fears the growing volume of AI-driven data traffic could overwhelm mobile network capacity, bottlenecking innovation and reducing the effectiveness of AI applications. Without sufficient full-power wireless spectrum, particularly in the mid-band range, Accenture warns networks will struggle to maintain the performance needed for seamless AI operations. The drag on AI development from limited wireless bandwidth contributes to an overall expected loss of $1.4 trillion in potential GDP by 2035 due to a lack of available spectrum.

Wireless networks will struggle to sustain required bandwidth levels without additional spectrum, ultimately leaving consumers at risk of losing access to AI applications that otherwise have the potential to transform daily life.Accenture, The Looming Spectrum Crisis

It is still early days in the AI revolution, and analysts are still trying to gauge the impact AI will have on mobile networks. Initial concerns have focused on uplink capacity, as networks were previously designed to handle download-heavy mobile video. Mobile industry analyst Joe Madden expects "AI video analytics will drive the next 'killer app' for 5G devices," and sees AI applications driving "uplink traffic 20% higher than it would be with existing applications."

Madden rightly calls on "Congress, the Administration, and the FCC to work together to clear and license spectrum for more network capacity." CSIS recently echoed this call, pointing to the need for spectrum auctions to support the digital infrastructure essential to U.S. leadership in AI. Accenture agrees, claiming "there is no time to waste on crucial spectrum policy," adding that, "U.S. policymakers should license additional, full-powered commercial mid-band spectrum to prevent detrimental effects to enterprises, consumers, and the country."

The availability of additional mid-band spectrum will enhance network capacity, ensuring AI applications function smoothly without congestion or service degradation. By accelerating spectrum auctions and reallocation efforts, policymakers can help ensure that mobile networks are equipped to handle the growing AI-driven data surge.

To fully unlock AI's potential, the mobile industry must have access to sufficient spectrum resources. Expanding mid-band availability will not only alleviate network strain but also drive continued AI innovation, supporting economic growth and technological advancement in the digital era.

Read the Accenture Report