Skip to content
← All Articles
2026-04-20 · AI

The AI Prosperity Gap: The Surprising U-Shape of AI Adoption by Income

Census HTOPS data reveals a surprising pattern: the $25K-$35K bracket uses AI more (29%) than those earning $150K+ (25%). The income story is more complex than expected.

The 2026 Census HTOPS survey reveals something unexpected about AI adoption and income: the relationship isn't linear — it's U-shaped. The $25K–$35K bracket uses AI at 29.28%, the highest of any income group, while those earning $150K+ come in at just 25.08%. The simple narrative of "rich people use AI more" doesn't hold up.

The Income × AI Cross-Tab

The full income breakdown tells a nuanced story:

- Under $25K: 18.83%

- $25K–$35K: 29.28% — the highest of any bracket

- $35K–$50K: 22.11%

- $50K–$75K: 22.96%

- $75K–$100K: 24.37%

- $100K–$150K: 26.89%

- $150K+: 25.08%

The ratio between the highest ($150K+) and lowest (under $25K) brackets is just 1.33× — far from a dramatic divide. And the $25K–$35K bracket actually exceeds the top earners by 4 percentage points.

The U-Shape Story

Why do lower-middle-income Americans lead AI adoption? Several theories:

Practical necessity: Americans earning $25K–$35K may use free AI tools like ChatGPT for practical tasks — price comparison, budget management, job searching, navigating bureaucracy. When money is tight, a free productivity tool is more valuable.

Digital natives in entry-level jobs: This bracket skews younger — digital natives comfortable with AI tools despite modest incomes.

The middle-class plateau: The $35K–$75K range shows the lowest adoption (22–23%), suggesting a "comfortable enough" zone where the urgency to adopt new tools is lower.

The Real Divide

The under-$25K bracket at 18.83% does show a meaningful gap — but it's a gap of access and awareness, not a 3× chasm. The true digital divide may be less about income and more about context: whether your workplace introduces AI, whether your social circle uses it, and whether you see it as solving a real problem in your life.

Education Matters More Than Income

Education correlates more strongly with AI adoption than income:

- Bachelor's degree: 28.62%

- Associate degree: 28.31%

- Graduate degree: 27.59%

- High school diploma: 22.78%

The education gap is wider and more consistent than the income gap, suggesting that familiarity with information tools — not purchasing power — drives adoption.

What This Means

The U-shaped pattern is actually encouraging. It suggests that free AI tools are already reaching lower-income Americans who find them useful. The policy implication: expanding awareness and digital literacy may matter more than subsidizing access. The tools are already free — the gap is knowing they exist and how to use them.

*Correction (April 27, 2026): This article originally stated a 3× gap between high and low-income AI usage. The actual ratio is 1.33× ($150K+: 25.08% vs under $25K: 18.83%). The article has been rewritten to reflect the more accurate — and more interesting — U-shaped adoption pattern. See our corrections page for details.*

*See who is lobbying to regulate AI at theailobby.com. Explore which jobs face the most AI exposure at aiexposure.org.*

Data source: U.S. Census Bureau HTOPS, March 2026.

More Analysis

Data from U.S. Census Bureau Household Trends and Outlook Pulse Survey (HTOPS), March 2026.