The Digital Divide in 2026: Which Americans Still Lack Broadband Access

The Divide That Persists

The term “digital divide” entered the American policy lexicon in the mid-1990s, when dial-up internet was still a novelty. Three decades later, the divide has not closed — it has shifted. High-speed broadband is no longer a luxury. It is the infrastructure through which Americans access healthcare, education, employment, government services, and economic opportunity. Yet approximately 19 million Americans still cannot purchase broadband service meeting the FCC's current 100/20 Mbps benchmark at any price.

This report maps the digital divide at a granularity that earlier data could not support. Using address-level records from the FCC's Broadband Data Collection (BDC), aggregated through our H3 hexagonal grid at approximately 500-meter resolution, we identify exactly which communities remain unserved, quantify the rural-urban gap across every state, and assess the demographic dimensions of disconnection.

The data tells a story that is both familiar and urgent. The communities with the least broadband access are overwhelmingly rural, disproportionately low-income, and often home to Indigenous and minority populations. The geography of disconnection maps closely onto the geography of poverty and historical underinvestment.

Source: FCC Broadband Data Collection, 2026

Defining “Unserved” and “Underserved”

The federal government uses two critical thresholds in its broadband policy framework. An address is classified as unserved if it lacks access to any provider offering at least 25/3 Mbps (25 Mbps download, 3 Mbps upload). It is classified as underserved if it lacks access to 100/20 Mbps service. These definitions matter enormously because they determine eligibility for BEAD funding and other federal subsidy programs.

By the unserved threshold, approximately 7.2 million Americans have no wired broadband access whatsoever — their only option is satellite service (primarily Starlink or HughesNet) or cellular hotspots. By the underserved threshold, the number jumps to 19 million, including millions of households stuck on legacy DSL connections that technically provide service but at speeds inadequate for modern use.

It is worth noting that the 100/20 benchmark itself is already falling behind. The FCC has signaled it may revisit this threshold as symmetric gigabit service becomes the deployment standard for new fiber construction. If the benchmark were raised to 100/100 Mbps (symmetric), the number of underserved Americans would increase substantially, as most cable and DSL connections cannot deliver 100 Mbps upload speeds.

The Rural-Urban Gap: By the Numbers

The most fundamental dimension of the digital divide is geographic. Urban addresses in the United States average 4.1 wired broadband providers offering 100/20 Mbps or faster service. Rural addresses average just 1.4. This nearly threefold gap drives differences in speed, price, and service quality that affect every aspect of daily life in underconnected communities.

The states with the widest urban-rural gaps reveal the structural nature of the problem:

In Mississippi, for example, 81% of urban addresses have broadband access, a number that drops to just 42% in rural areas — a 39-percentage-point gap. The Delta region of Mississippi contains some of the most underserved ZIP codes in the country, with multiple communities where zero wired broadband providers report service availability.

Source: InternetProviders.ai analysis of FCC BDC data, March 2026

America's Most Underserved Regions

While the digital divide exists in every state, our analysis identifies six regions where the concentration of underserved addresses is most severe:

Tribal lands represent the most extreme expression of the digital divide. On the Navajo Nation, which spans parts of Arizona, New Mexico, and Utah, broadband coverage at the 100/20 standard is estimated at roughly 30%. Many homes lack not only broadband but also electricity and running water — infrastructure deficits that make broadband deployment exponentially more challenging and expensive.

The permitting process for broadband construction on tribal lands adds another layer of complexity. Federal environmental reviews, tribal sovereignty considerations, and cultural resource protections can extend deployment timelines by 12 to 24 months beyond what comparable off-reservation construction requires. The BEAD program includes specific tribal set-asides, but the administrative and logistical barriers remain formidable.

In the Texas border region specifically, colonias and small unincorporated communities face some of the sharpest connectivity gaps in the nation. Residents searching for available providers in these areas can use InternetNearMe.ai to check address-level broadband availability across Texas communities.

Demographic Dimensions of Disconnection

The digital divide is not only geographic — it is demographic. Cross-referencing FCC availability data with Census Bureau demographic data reveals patterns that reinforce existing inequalities:

• Income: ZIP codes where the median household income is below $35,000 have an average broadband coverage rate of 68%, compared to 92% for ZIP codes with median incomes above $75,000. Even where broadband is technically available, affordability remains a barrier in low-income communities.
• Age: Counties with a higher share of residents aged 65 and older tend to have lower broadband adoption rates, even when infrastructure is available. This reflects both affordability concerns on fixed incomes and digital literacy gaps.
• Race and ethnicity: Predominantly Black counties in the rural South and predominantly Native American areas in the West have broadband coverage rates 15 to 25 percentage points below the national average. These patterns reflect decades of infrastructure underinvestment that predates broadband.
• Education: Areas with lower educational attainment correlate with lower broadband availability, creating a feedback loop: limited connectivity constrains educational and economic opportunity, which in turn limits the community's ability to attract investment in infrastructure.

These demographic patterns underscore that closing the digital divide requires more than infrastructure deployment alone. Affordability programs, digital literacy initiatives, and device access programs are all necessary components of a comprehensive strategy.

Source: U.S. Census Bureau American Community Survey, 2024

Provider Coverage Gaps: Who Serves Where

The major national ISPs have fundamentally different geographic footprints, and understanding those footprints is essential to understanding the divide:

• AT&T Fiber has aggressively expanded its fiber network to approximately 26 million locations, but its coverage is concentrated in suburban and urban markets across its 21-state footprint. Rural areas within AT&T's territory often remain on legacy DSL or have been effectively abandoned. See our AT&T coverage analysis.
• Xfinity (Comcast) covers approximately 62 million homes, almost entirely in suburban and urban areas. Comcast has publicly stated it does not pursue rural buildouts without government subsidy. See our Xfinity coverage page.
• T-Mobile Home Internet has been the single most significant expansion of broadband access in rural areas over the past two years, reaching an estimated 55% of U.S. addresses. However, fixed wireless performance varies dramatically based on tower proximity and congestion. See our T-Mobile analysis.
• Starlink provides satellite coverage to virtually 100% of U.S. addresses, but at higher cost ($120/month) and with latency that makes it unsuitable for some applications. For the most remote areas, it remains the only viable option. See our Starlink review.

The bottom line: no single provider is solving the digital divide. The most underserved areas require either government-subsidized buildouts (via BEAD, RDOF, or state programs) or a combination of technologies — fiber for trunk infrastructure, fixed wireless for middle-mile distribution, and satellite for the most remote last-mile connections.

Will BEAD Close the Divide?

The $42.45 billion Broadband Equity, Access, and Deployment (BEAD) program is the largest federal investment ever directed at closing the digital divide. Every state has received its allocation, and as of early 2026, several states have begun awarding subgrants for construction. The question is whether the funding is sufficient and whether it will reach the communities that need it most.

Our analysis suggests the answer is cautiously optimistic but incomplete. BEAD funding is sufficient to extend fiber or fixed wireless service to the vast majority of unserved and underserved locations — estimated at 8 to 10 million addresses — but the timeline is long. Most states will not complete their BEAD-funded buildouts until 2028 or 2029. Construction in challenging terrain (Appalachia, Alaska, tribal lands) will take even longer.

There are also structural risks. Workforce shortages in fiber construction could delay projects. Material costs (particularly fiber optic cable and electronics) have risen. And the program's requirement that funded networks deliver at least 100/20 Mbps could become outdated by the time construction is complete, potentially leaving newly connected communities with infrastructure that is already below the standard of what urban Americans enjoy.

For real-time tracking of BEAD allocations and deployment progress, see our BEAD Funding Tracker.

Source: NTIA BEAD Program, 2026

Methodology

This report is based on InternetProviders.ai's analysis of the FCC's Broadband Data Collection filings covering 13.1 million availability records across all 50 states and the District of Columbia. We aggregate address-level data using H3 hexagonal grids at approximately 500-meter resolution, covering 10,105 cities and 30,501 ZIP codes.

Urban and rural classifications follow the Census Bureau's definitions based on population density thresholds. Demographic overlay data is drawn from the 2020 decennial census and the 2022 American Community Survey 5-year estimates.

Coverage percentages refer to the share of residential addresses with access to at least one provider delivering the specified speed threshold (100/20 Mbps unless otherwise noted). Provider counts exclude satellite-only providers. All data reflects the most recent BDC filing available as of March 2026.

Full methodology details are available on our methodology page. This research is published under a CC BY 4.0 license.

Source: InternetProviders.ai Methodology