Georgetown Law Universal Broadband Summit Report
An ISOC LIVE Summary
Date: July 23, 2026
Author: Stephanie Weiner — Federal Alumni Fellow, Georgetown University
The Georgetown Law Institute for Technology Law & Policy convened the Universal Broadband Summit on 21 May 2026 to identify the most significant remaining barriers to universal broadband in the United States. Participants included state broadband officials, industry associations, local government organizations, union representatives, academics, and civil society. The summit was deliberately future-focused: rather than relitigating how the country arrived here, participants took the current legal and programmatic landscape as a starting point and asked where gaps persist, why they have proven durable, and what states and stakeholders can do now. Conducted under the Chatham House Rule, the report is not an attributable record of the convening but a practical playbook organized around four challenges, each examined through root causes, replicable models, and metrics.
The report is candid about its own boundaries. It takes current law and the current federal framework as fixed, leaving aside questions of legality, legislative solutions, and Universal Service Fund reform, and treats the availability of remaining federal funding as uncertain. The summit included no representatives from the FCC, NTIA, the Rural Utilities Service, the White House, or Congress, so the findings reflect those implementing federal broadband programs rather than those designing them. Provider perspectives came largely through trade associations; satellite providers and fixed-mobile convergence were not fully examined. Participation does not imply endorsement of the findings.
Universal Broadband as Three Complementary Goals
The report argues that universal broadband requires more than network construction, and defines three components:
Availability: a high-speed network reaches the home.
Adoption: the household subscribes to it.
Utilization: people have the devices and skills to use it effectively.
Drawing on the FCC’s Internet Access Services: Status as of June 30, 2025 (published May 2026), the report finds that of approximately 116.4 million broadband serviceable locations, 111.2 million — about 95.5% — have access to fixed terrestrial service at 100/20 Mbps, leaving a 5.2 million-location availability gap concentrated in rural, Tribal, and other high-cost areas. Two caveats complicate that figure. Because the data are provider-reported, and a location counts as served if an ISP could provide service on request rather than whether a connection was actually deployed, the true gap may be larger. And the gap depends on where the goalpost sits: at roughly gigabit-class service — 940 Mbps down, 500 Mbps up — 47.5 million locations, nearly 41% of all BSLs, lack qualifying service.
Adoption tells a starker story. Of 130.6 million U.S. households, 93.1 million subscribe at 100/20 Mbps or above — 71.3% — leaving 37.5 million non-adopting households, more than a quarter of the country, with rates significantly lower among rural, Tribal, and low-income populations. The report frames adoption as a first-order concern for two reasons: connectivity’s benefits flow only from actually being online, and networks without enough subscribers to cover their costs will fail, taking public investment with them. Adoption is therefore not only an equity goal but the demand-side financial foundation that makes deployment durable. BEAD’s $42.45 billion has been directed at availability rather than adoption.
Challenge 1: Permitting Reform
Permitting was identified by more summit participants than any other issue as their top-priority challenge. The discussion builds on a prior Georgetown permitting summit held in 2024, memorialized in the Permitting Success Report published by the Benton Institute for Broadband & Society.
Three root causes emerged: a structural gap between permit volume and the finite capacity of processing offices — felt most acutely in rural jurisdictions least able to staff up for a temporary surge, and extending to the 811 “call before you dig” locate systems construction depends on; paper-based processes alongside fragmented, undigitized records of existing underground and aerial infrastructure; and breakdowns in coordination, communication, and trust between providers and permitting authorities, with Tribal consultation carrying its own legal and sovereign weight.
The models are operational rather than regulatory:
Surge support sends capacity where demand spikes — the Mesa, Arizona model, in which ISPs reimburse the city for third-party consultants working inside the permitting office under city standards; state-funded consultant pools routed through the broadband office to simplify procurement for smaller jurisdictions; regional “tiger teams” of experts on recurring obstacles like poles, railroads, and easements; and direct funding to permitting authorities and 811 systems.
Digitization addresses both the process — San José’s SJePlans portal paired with GIS layers, Rancho Cordova’s enterprise permitting system, Illinois’s and Michigan’s state-level platforms, NTIA’s ESAPTT for environmental screening — and the underlying data, through as-built reporting at project closeout and GIS platforms like Johnson County, Kansas’s AIMS. Participants flagged that a “digital twin” of underground infrastructure raises security concerns and data sovereignty concerns on Tribal lands, and that modernization may arrive too late to affect BEAD’s construction window.
Permitting playbooks are the least funding-dependent model: collaboratively developed guides describing what a productive first conversation covers and what each side brings, ideally created by a state broadband office in partnership with local governments rather than handed down. BEAD now requires states to convene permitting roundtables including federal, state, local, and Tribal representatives. Arkansas released a 75-page toolkit and runs a bi-monthly roundtable; Illinois developed a BEAD starter guide; Minnesota produced a flowchart allocating responsibilities among provider, state, and NTIA.
Measurement is difficult because success in permitting is invisible — when the process works, deployment simply proceeds — and because units are not standardized across jurisdictions. Proposed metrics pair throughput and timeliness (time to issuance and, separately, time to closeout) with accountability measures tying public dollars to outcomes, efficiency gains for providers and authorities, and damage-prevention outcomes benchmarked against the Common Ground Alliance’s DIRT dashboard.
Challenge 2: Demand-Side Initiatives
Demand-side initiatives, including affordability and adoption, emerged as the most frequently identified top-five priority in the pre-event survey.
The report separates two problems that are often conflated. The first is market concentration: where a household has only one provider, the absence of competition keeps prices high and quality low regardless of income. FCC data as of June 2025 show that at 100/20 Mbps, one-quarter of BSLs had one or zero fixed terrestrial providers and well over half had no more than two. A Federal Reserve Bank of New York study found the least-affordable city charged among the highest prices for some of the lowest-quality service, attributing the pattern to the absence of a competitive marketplace. The second is affordability: even a competitive price can exceed what a low-income household can pay. The same study found low- and moderate-income communities spend roughly 2.43% of median household income on broadband, nearly five times the 0.51% share in wealthier areas and above the FCC’s 2% benchmark.
ACP’s lapse in May 2024 removed the country’s largest affordability tool, having reportedly helped over 20 million households. Lifeline, the standing federal program, offers $9.25 monthly, reaches a small share of eligible households, and draws on a shrinking revenue base. Further root causes include the device barrier — roughly 1 in 7 households nationally, and as many as 1 in 4 in some areas, lack a desktop or laptop, and research by Gonzales and Zhang finds computer access a stronger predictor of beneficial internet use than in-home service — along with evolving skills demands (participants noted the baseline has shifted from operating a mouse to navigating AI tools and recognizing sophisticated scams), trust and online-safety concerns that lead some households to stay offline, and a pervasive data gap: no comprehensive public record exists of what households actually pay.
Five families of models follow:
Competition. Lower barriers to entry. Colorado’s repeal of its 2005 municipal referendum requirement has been followed by more than 120 communities moving to offer or support local options, one providing service at $40 per month, or $15 for low-income households.
State-funded subsidies. New Mexico’s Senate Bill 152, signed February 2026, establishes the Low-Income Telecommunications Assistance Program, providing up to $30 per month for households earning $45,000 or less — funded through the State Rural Universal Service Fund rather than general appropriations, and structured to scale from $10 million in year one to $45 million annually.
Low-cost plan requirements. New York’s Affordable Broadband Act, upheld by the Second Circuit with certiorari denied, requires providers with at least 20,000 in-state customers to offer eligible households $15 service at 25 Mbps or $20 at 200 Mbps. Participants were candid about replication difficulty, noting New York’s unusually concentrated market. More consequentially, NTIA’s June 2025 BEAD Restructuring Policy Notice bars states from setting low-cost service option rates, and NTIA’s general terms require states to agree not to enforce any state law regulating the rates, terms, or conditions of a BEAD subgrantee’s service statewide — including for households outside the subgrantee’s BEAD footprint. That asserted condition has not been tested in court, and some question the agency’s authority to condition grants on non-enforcement of duly enacted state law. At least one state set aside a New York-style bill in 2025 after concluding it could place its BEAD allocation at risk.
Devices and digital skills. Oregon’s 2025 Lifeline reform pairs an increased subsidy with a one-time $100 device benefit. ACP’s comparable device discount went largely unused because most providers declined to offer it — an argument for routing device support through institutions equipped to stock and fulfill hardware. WIOA funds can support digital-skills instruction, and participants stressed delivery through trusted local institutions, with libraries serving a triple role as device lenders, public access points, and hands-on help.
Price and adoption transparency. A comparatively low-cost model, with consumer-protection measures advancing on a bipartisan basis in Virginia and Arkansas.
Metrics are organized by program type. The report is explicit that the decision to measure must be made when a program is designed rather than when results come due, and that a well-built metrics framework helps close the demand-side data gap even as it tracks outcomes.
Challenge 3: Workforce Development and Scaling
Participants agreed that BEAD, private investment, and rapid AI-infrastructure expansion are creating a historic surge in labor demand, with estimates suggesting tens of thousands of new workers will be required to meet deployment deadlines.
Three root causes emerged:
Communication and data gaps. Many major ISPs have not raised workforce shortages as urgent, because construction is largely performed by contractors and subcontractors who experience the most acute shortages but have little direct engagement with state broadband offices. No comprehensive source measures workforce needs specific to broadband deployment; federal labor data rely on job classifications that map poorly onto the work.
Pipeline and job quality. A 2024 study for the Fiber Broadband Association and Power & Communication Contractors Association estimated the industry will need roughly 28,000 new construction jobs and 30,000 new technician jobs over the next decade, plus 56,000 and 64,000 replacement workers respectively. Pew reports roughly 1 in 5 telecommunications workers are now over 55, with about 120,000 replacements needed by 2032 independent of new demand. Hyperscaler data-center construction competes for the same electricians, technicians, and construction workers, with deeper pockets, in-house training, and no BEAD constraints. Notably, labor representatives dissented from the pipeline framing, maintaining the binding constraint is a shortage of quality jobs — stagnant real wages and eroded career ladders as providers shifted from in-house workforces to contractors — and that stable, well-paying jobs with advancement pathways would draw a substantial pool of qualified workers, including recently laid-off legacy telecom workers. The report treats the two framings as compatible: the pipeline is inadequate, and job quality determines whether trained workers stay.
Inconsistent, delayed funding. Training dollars often arrive after construction demand peaks, leaving programs launched in anticipation of promised funds without resources to continue.
The models span the pipeline:
Sector partnerships convening industry, government, education, and labor — Ohio’s Strengthening Ohio’s Broadband & 5G Workforce strategy, anchored by Ohio State, brought together more than two dozen partners. Participants stressed that employers must help specify training from the start, and that partnerships typically require Governor’s-office support because workforce policy sits outside the broadband office.
Recruitment and awareness, including NTCA’s BOLD Toolkit for provider–K-12 collaboration, Ohio’s High School Tech Internship Program, and mobile demonstration units. Participants urged an “all-of-the-above” strategy reaching non-traditional labor pools, paired with wraparound supports like childcare and transportation.
Credentialed training programs, including Louisiana’s ConnectLA partnership with the state community college system (an initial $10 million investment; more than 850 completions), NTCA’s Broadband Academy at Northwood Tech, the FBA’s OpTIC Path certification (roughly 70 institutions across 40 states), the Arkansas Fiber Academy, New Mexico’s free monthly certification classes, and the Cherokee Nation’s Fiber Technician Program.
Registered apprenticeships and intermediary models, where a trade association handles compliance so small employers can participate — notably TIRAP, led by the Wireless Infrastructure Association, with competency-based standards across 16 occupations and over 100 employers. Union-affiliated pathways run through CWA and IBEW.
Throughout, participants emphasized that credentials should be stackable and transferable across employers and sectors. Measurement is unusually hard here: no accepted alternative to the inadequate “day-one hire” rate has emerged, longitudinal tracking has proven difficult to implement, and the employer-centric framework underlying most workforce measurement fits poorly with a goal of worker mobility. Proposed metrics span completion and credentialing, placement and employer demand, sustained and quality employment, and value-added earnings.
Challenge 4: Default Risk and Connecting Remaining Locations
Some BEAD-funded projects will fail, and some locations will remain unconnected when Final Proposal awards conclude. Several stakeholders have proposed reserving part of the roughly $22 billion in remaining BEAD funds; the Vernonburg Group has recommended a $7.6 billion “rainy day” fund covering RDOF defaults, uncaptured locations, and BEAD cost overruns. Neither the statute nor NTIA guidance provides a ready mechanism.
Root causes:
Single-round program design. BEAD funds one round of deployment with no statutory mechanism to “true up” the eligible-locations list or redeploy clawed-back funds.
Maps lag ground truth. One analysis identifies approximately 1.1 million unserved and underserved locations not funded by any federal program; CostQuest separately estimates roughly 2.1 million BSLs omitted from Final Proposals will remain unserved or underserved.
Legitimate cost increases. Rising pole attachment fees and data-center-driven fiber demand can erase project margins. RDOF, the closest analogue, has seen defaults on roughly 37% of awarded locations — more than 1.9 million locations and nearly $3.3 billion.
Uncertain scale and timing. Neither the number nor the timing of defaults can be predicted, arguing for a flexible framework.
Rather than cataloging models, participants built a three-phase framework:
Phase 1 — Prevent. Structure grant agreements with risk assessments, milestone monitoring, and early-warning mechanisms; calibrate compliance to a “Goldilocks” standard rigorous enough to catch problems early but not so burdensome that compliance itself causes default; monitor proactively and predictively as “enhanced technical assistance” rather than punitive oversight; provide state-funded technical assistance; and hold providers responsible for their own compliance and project-management capacity.
Phase 2 — Preserve. Distinguish legitimate cost overruns from manufactured ones — a tripling of attachment fees may signal rent-seeking, while data-center-driven fiber costs reflect genuine programmatic increases. Participants favored giving states flexibility, subject to NTIA reasonableness review, to supplement awards for verified overruns as budget modifications rather than treating them as noncompliance; endorsed a “stick with the provider you have” default; and urged states to formalize a community of practice for cost benchmarking, polling peer states in near-real time when a multi-state provider makes a particular claim.
Phase 3 — Connect. Address remaining locations — both defaulted BEAD-eligible locations and non-BEAD locations omitted from Final Proposals, likely the highest-cost and hardest to reach. Three pathway choices carry tradeoffs: negotiation with prior bidders versus competitive rebid; immediate response as defaults arise versus a set cleanup date; and rolling awards versus batched awards that may attract larger providers at the cost of delay. Participants supported NTIA establishing outer bounds and authorizing both options, leaving the choice to states.
Metrics pair leading indicators — milestone progress and financial early-warning signals intended to surface likely defaults 12 to 24 months ahead — with outcome measures of defaults, funds at risk, at-risk projects preserved, and locations actually connected, tracked against a denominator that should decline toward zero.
Overall Findings
No single barrier stands between the country and universal broadband, and no single solution clears the path. The report offers replicable models, practical frameworks, and lessons drawn from states already doing this work, with emphasis throughout on measurable outcomes and continuous evaluation — so that progress toward universal connectivity can be demonstrated rather than merely asserted.
RESOURCES
Universal Broadband Summit Report — Georgetown Law Tech Institute, July 2026, CC BY-NC 4.0
Stephanie Weiner — author, Federal Alumni Fellow and former NTIA Chief Counsel
Georgetown Law Institute for Technology Law & Policy — convener
Internet Access Services: Status as of June 30, 2025 — FCC report (May 2026) underlying the availability and adoption figures
Permitting Success — Drew Garner, Benton Institute (Sept 2024), the prior summit’s report
Broadband Affordability — Federal Reserve Bank of New York study on low- and moderate-income pricing
Achieving Affordability — Jordan Arnold, Benton Institute, on state affordability strategies
BEAD Restructuring Policy Notice — NTIA (June 2025), source of the low-cost service option rate restriction
NTIA Encourages ISPs to Ensure State BEAD Contracts Comply With Agency Priorities — Drew Garner on the unresolved rate-regulation questions
An Updated Estimate on the Size of a BEAD Rainy Day Fund — Vernonburg Group’s $7.6B proposal for defaults and unfunded locations