Broadband Is Not Enough: The Missing Infrastructure Between Communities and AI

For years, broadband policy has focused on one essential goal: connecting every home, business, school, farm, and community anchor institution to affordable high-speed internet. That work remains critical. But the rise of artificial intelligence is forcing a new question that many states, communities, and local network operators have not yet had to confront: once users are connected, where does their traffic actually go?

AI will not run on last-mile connectivity alone. Real-time inference, AI agents, telehealth, cybersecurity, precision agriculture, advanced manufacturing, education, and public-sector AI services will depend on fast, predictable, resilient connections between users, networks, cloud platforms, data centers, and compute resources. If traffic from a community must travel hundreds of miles to reach the networks and compute platforms it depends on, physical distance, inefficient routing, packet loss, and congestion can become hidden barriers to performance, adoption, and economic competitiveness.

This session will help state broadband leaders, local ISPs, policymakers, and community stakeholders understand the missing layer of infrastructure between broadband access and AI readiness: the regional places where networks meet, exchange traffic, connect to cloud and compute resources, and keep data closer to the communities it serves.

Panelists will explain why Internet Exchange Points (IXPs), regional transport, neocloud connectivity, and compute-adjacent interconnection are becoming essential building blocks for AI-ready regions. The discussion will challenge the assumption that distant data centers, national carriers, and major cloud platforms will simply “make it work” for every community—and will explore why states that fail to develop regional interconnection infrastructure may risk falling behind in the next wave of the digital economy.