Small-business technology adoption is measurable, but adoption data does not remove the need for people who check sources, exceptions, and customer context. A buyer may compare a virtual assistant with software because both can handle parts of an administrative task. Census BTOS data reported that overall U.S. business AI use hovered between 17% and 20% from December 2025 through May 3, 2026, while 20% to 23% expected to use AI in the next six months. Those figures describe business use, not job replacement or output quality.
Small-business VA adoption and demand signals
The BTOS asks businesses about recent and expected use. It is a high-frequency, nationally representative survey of employer businesses, with experimental AI measures. It does not classify every tool, measure the quality of an AI output, or determine whether a person’s work will disappear.
Key stats
Verified August 11, 2026Key takeaways
- AI use was substantial but far from universal in the Census business measure.
- A technology-use percentage does not say whether a business has reliable data, review capacity, or a safe process.
- Human support remains relevant where a task needs context, follow-up, source checking, or a decision owner.
Where human support still matters
Many administrative tasks combine structured steps with unstructured judgment. A person can check whether a source answers the question, notice a missing customer detail, ask for clarification, or route an unusual case. These are service outcomes, not simply keystrokes.
Compare tools and people on the same output
Ask each option to produce the same deliverable from the same safe sample. Score factual accuracy, completeness, turnaround expectations without a guarantee, exception handling, and manager effort. Keep the result tied to the business task rather than a broad claim about AI or people.
Consolidated statistics
| Measure | Finding | Limit or use |
|---|---|---|
| BTOS measure | Reported range | What it leaves open |
| Recent AI use | 17% to 20% | Which tools and tasks? |
| Expected AI use | 20% to 23% | Whether adoption works |
| Sample design | 1.2 million businesses | Survey response limits |
For a related buyer context, read the research support. For a second comparison, see the productivity research.
Methodology and limits
Facts were checked against the linked public sources on August 11, 2026. Findings include the unit, period, population, and source context available in each register. Survey and incident datasets can be revised or sampled, and none predicts an individual provider, worker, or buyer outcome.
Frequently asked questions
Does AI adoption mean a business no longer needs assistants?
No. The Census measure reports business use, not job replacement. People may still be needed to check evidence and handle context.
Is the BTOS AI measure a productivity study?
No. It is a survey measure of use and expectations. It does not prove that a tool improves a particular workflow.
How should a buyer test the options?
Use the same safe sample and score the completed output, missing information, and review effort.
Sources
10 visible sources, verified August 11, 2026
- Census AI use at U.S. businesses
- Census BTOS data downloads
- Census BTOS about
- Census BTOS help
- Census BTOS methodology
- NIST AI risk management framework
- NIST small business cybersecurity
- SBA technology resources
- BLS productivity program
- FTC business guidance on AI
A practical next step
Use the evidence as a starting point, then compare one defined role with a safe work sample and a named review owner. A free consultation can help scope that comparison without treating a public statistic as a promise.
