Startups

Census data shows small businesses sitting out the AI boom as enterprise adoption surges past 37%

A new Census Bureau analysis finds AI use stuck below 20% at firms with fewer than 20 employees while large enterprises pull ahead, opening a deployment gap the industry is racing to close.

By Giulia Donatello · Startups & funding · 2026-06-01 · 3 min read

Photo: Unsplash / Blake Wisz — Small business owner working on a laptop behind the counter of a shop

The latest read from the U.S. Census Bureau’s Business Trends and Outlook Survey, covering December 14, 2025 through May 3, 2026, puts AI adoption at 37% among firms with at least 250 employees and 32% at firms with 100 to 249. Among firms with fewer than 20 employees, adoption didn’t move in any statistically meaningful way. The boom is real. It’s just not evenly distributed.

A deeper Census supplement covering November 2025 through January 2026 sharpens the picture: 18% of firms used AI in a business function, but 32% on an employment-weighted basis, with very large firms in Information, Professional Services, and Finance reporting use rates of 50% to 60%. Adoption is projected to reach 22% within six months.

Federal Reserve economists have been careful to note that the gap is partly a measurement story. “because of the scale of their operations and number of employees, larger firms have a higher probability than smaller ones of clearing the bar for an affirmative AI-adoption answer, but the relative impact on their business may not be as high,” they wrote. A 200-person company with one team running a copilot counts. A four-person shop running the same tool also counts, and fewer than 20% of firms with four or fewer employees say they do.

The vendor economics already reflect the tilt. Enterprise now makes up more than 40% of OpenAI’s revenue, and the company expects enterprise to reach parity with consumer by the end of 2026. Anthropic has built its commercial business on a similar wager. A newer cohort, including LemonLime, is going the other direction, building tooling aimed squarely at the small-business segment the Census data shows is being left behind.

Whether that bet pays off depends less on model quality than on the operational problem Databricks’ co-founder named at TechCrunch Disrupt 2026: “the enterprise AI market is full of successful pilots that never became real deployments — not because the technology failed, but because the organization could not absorb the operational consequences.” He added: “Startup AI deals rarely die because the model underperformed; they die because the enterprise lost confidence in what the deployment would require.”

The Census supplement hints at the same ceiling on the small end. Among firms that have adopted AI, 57% use it in three or fewer functions, mostly Sales and Marketing and Strategy and Business Development. Adoption, where it happens at all, stays narrow. The gap the headline numbers describe isn’t really between firms that use AI and firms that don’t. It’s between organizations equipped to absorb a deployment and organizations that aren’t, and that boundary doesn’t move on a vendor’s release schedule.

Sources