AI is no longer confined to data centers or research labs. It is increasingly mobile, real‑time and woven into everyday life — from navigation apps and virtual assistants to industrial automation and connected vehicles. Yet as policymakers debate the future of AI, one foundational reality is often overlooked: AI only works if a robust digital infrastructure that supports it is in place.
New research from the Wireless Infrastructure Association (WIA) makes this clear. Our recent survey of U.S. consumers found that 74% of U.S. adults now use AI applications, and a majority do so on the go — relying on mobile broadband networks to access intelligence in real time. Already, AI traffic accounts for more than 4% of total U.S. wireless network traffic, representing nearly $3 billion annually in network investment and operating costs, according to WIA’s new analysis — and that share is rising rapidly as AI adoption accelerates.
This data underscores a simple truth: America’s AI future depends on America’s wireless infrastructure. To achieve critical national objectives for leadership in AI innovation, productivity and economic growth, the U.S. must ensure that digital infrastructure and spectrum policies are aligned with the scale and speed of what’s coming next.
Infrastructure is the unsung AI enabler
Wireless networks are already doing the heavy lifting for America’s connectivity needs. According to WIA’s Infrastructure by the Numbers research, the U.S. wireless industry invested more than $65 billion in wireless infrastructure in 2025 alone, supporting nearly 158,500 purpose-built cell towers, hundreds of thousands of macro and small cells and millions of square miles of coverage nationwide. This infrastructure doesn’t just connect phones; it supports the data flows that power modern services and applications.
As AI becomes more embedded in daily life, demands on networks will grow exponentially — not just in city centers, but in suburban, rural and indoor environments where people live and work. Network demands from AI will come from more than just AI on the network from consumers. AI-enabled IoT devices, as well as AI in the network via the tools used to manage and optimize the mobile network itself, including AI-RAN, will also require robust connectivity.
Additionally, wireless infrastructure will have a role to play in edge AI to enable applications like autonomous systems, immersive computing and AI‑powered industrial tools that require super low latency, high reliability and ubiquitous coverage. Due to these factors, network operators are expecting a massive wave of data demand rising from AI in the next few years.
“As we move into what I’d describe as the AI hypergrowth phase, we expect data consumption to accelerate by another 80% over the next five years,” AT&T CEO John Stankey said in an interview with Brunswick Review last year.
Permitting and spectrum policy must match the moment
Infrastructure investment alone isn’t enough. Spectrum is the fuel that powers wireless networks, and today’s spectrum pipeline is dangerously thin relative to future AI-driven demand. Policymakers must act with urgency to fill the pipeline using a forward‑looking spectrum strategy — one that makes sufficient licensed spectrum available, supports innovation across bands and avoids artificial constraints that slow deployment.
This is not a theoretical concern. AI is already reshaping traffic patterns and network economics. Without timely spectrum decisions and modernized infrastructure policies, the U.S. risks bottlenecks that could slow AI adoption, increase costs and undermine global competitiveness.
6G: The connectivity fabric for AI
Industry leaders are already looking ahead to what comes next. As Nvidia, one of the world’s leading AI companies, has emphasized, future wireless generations will be inseparable from AI itself.
At the Brooklyn 6G Summit, NVIDIA executive Ronnie Vasishta put it plainly: “6G really distributes AI to the entire population and enterprises. It’s the connectivity fabric for AI, and it cannot be underestimated how important that is.”
That insight should resonate in Washington D.C. If 6G is the fabric that connects AI everywhere — from cloud to edge to device — then the policy decisions we make today on infrastructure siting, spectrum availability and investment will determine whether the U.S. leads or lags in the next era of innovation.
AI policies that succeed in practice
America has led before by pairing technological innovation with smart, enabling policy. AI demands the same approach. That means:
- Accelerating infrastructure deployment through a national framework to streamline rules for predictable, proportionate and transparent permitting processes.
- Filling a robust spectrum pipeline that keeps pace with AI‑driven demand by ensuring full-powered licensed spectrum in critical 2.79, 4 and 7 GHz bands.
- Recognizing wireless infrastructure as essential AI infrastructure, not an afterthought.
AI doesn’t exist in a vacuum. It exists on and in networks — built, upgraded and operated every day by the wireless infrastructure industry. If we want AI policies that succeed in practice, not just on paper, we must start by ensuring the infrastructure that supports AI is ready for the future.
Patrick Halley is the President and CEO of the Wireless Infrastructure Association (WIA), which represents the businesses that build, develop, own and operate the nation’s wireless infrastructure and advocates for the widespread, responsible deployment of wireless infrastructure to enable connectivity everywhere.
Opinion pieces from industry experts, analysts or our editorial staff do not represent the opinions of Fierce Network.