Spectrum, 6G, and the Future of AI Leadership

As AI continues to proliferate from autonomous systems to industrial automation, so too will the demand for quick, reliable, low-latency connectivity. This evolution will place spectrum policy as a central pillar for AI advancement, underpinning the key infrastructure needed for these systems to deploy. The advancements of 6G technology—the sixth generation of mobile communication technology set to surpass 5G in speed, connectivity, and AI integration—will also accelerate the demand for broader spectrum. As 6G is being developed with AI integration, the capability of this technology will depend on access to midband spectrum, further integrating the importance of spectrum, AI, and telecommunications policy as interdependent. The United States will need to expand the spectrum pipeline to meet the demand of AI, 5G, and 6G, which is increasingly reliant on high-speed, low-latency connectivity for training, inference, and edge computing. Given this convergence of factors, spectrum policy is increasingly resembling AI policy, and the failure to adapt to the needs of the present moment ultimately risks U.S. competitiveness against China.

Lessons Learned from 5G Deployment

AI deployment will be closely tied to advances in wireless technology, especially 6G, making spectrum so essential. In a presential memorandum in December of 2025, President Donald Trump outlined ways in which 6G and next-generation wireless networks are essential to the rapid deployment of AI, in turn affecting the economic prosperity and national security of the United States. During the 5G rollout over the last decade, mid-band spectrum access was decisive in early deployment advantages. Where the United States lagged in some allocations, hampering U.S. companies’ competitiveness, China moved faster. The result was the growth of Chinese 5G infrastructure globally, with Huawei becoming one of the largest providers of 5G mobile technology. This reality continues to have long-term consequences for the United States, including concerns over cyber espionage and the access the Chinese Communist Party has to massive amounts of data on communications networks and critical infrastructure. The 6G buildout, by contrast, will have to move faster; with AI natively integrated into this architecture, fully realizing these capabilities will require a dramatic scaling of network capacity.

The Next Frontier for AI 

AI is also moving beyond cloud-based, text-driven tools such as chatbots, toward more latency-sensitive applications. These include autonomous systems like drones and robotics that have both critical civilian and military uses. Unlike text-driven tools, these systems will require near instant response times and necessarily move the compute closer to the user and device. This expected transition will need to consider the tradeoffs for cloud-based and on-device AI. Unlike cloud-based AI, device AI allows for increased speed and latency to the user, which is essential for autonomous systems. It is constrained, however, by compute power, battery life, and limited datasets. Whereas cloud-based AI struggles with latency, it has a greater capacity for larger datasets and more powerful computational abilities. This is where edge computing and the transition to 6G become essential. Edge computing emerged during the 5G cycle and moves compute closer to the user. But, unlike 5G, AI will be a native technology to 6G, meaning it will already be embedded in the foundational architecture, enabling larger datasets and faster processing times. 6G will underpin AI-enabled systems across military, industrial, and civilian domains.

Why Spectrum Matters for 6G’s Success

Latency-sensitive applications, like drones and industrial robots, need improved wireless infrastructure—deploying AI in the physical world requires increased mobile connectivity. As AI transitions to a reality in which people are using more vision-based and other latency-sensitive applications, it will lead to increased data moving along wireless networks. Latency, bandwidth, and reliability thus directly shape AI performance.

One of the most fundamental aspects of this wireless infrastructure is mid-band spectrum. The wireless industry today was built around narrow blocks of spectrum, and it is not sufficient for the advancements of 5G, 6G, and AI applications that require much larger swaths of spectrum. While the upper C-band auction scheduled for 2027 and forward progress on the 2.7 gigahertz (GHz) band can likely relieve some of the near-term pressure for capacity, the 4 GHz and 7 GHz bands will be key to provide the large blocks of contiguous spectrum needed for 6G At the same time, the current spectrum allocation process is slow and fragmented, with competing demands between government, including defense, and commercial users.

However, as the deployment and expansion of AI and 6G continue—with 6G expected to be deployed in 2030—the dual-use nature of these technologies and the growing technological competition between China and the United States increasingly reveals the blurry line between commercial and defense systems, making the allocation of mid-band spectrum a national security imperative. The future of AI advancement in the United States will require infrastructure buildout to support growing data demands and uses like autonomous systems. Without sufficient spectrum, powerful AI systems and the 6G wireless technologies utilizing them cannot be deployed in the real world at the necessary speed and scale. Fragmented or even delayed U.S. spectrum policy reforms could thus constrain domestic AI deployment, leading to competitors like China gaining the advantage in distributed AI systems and applications.

With 5G deployment and Huawei’s global position as a cautionary warning, the United States should waste no time in equipping U.S. firms with the essential spectrum necessary to dominate the AI and 6G technology race.

Taylar Rajic is an associate fellow with the Strategic Technologies Program at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Matt Pearl is director of the Strategic Technologies Program at CSIS.