82% of Small Businesses Now Use AI. Here Is What That Actually Means.

A new JPMorganChase Institute study published this month found that 82 percent of small business employers have adopted at least one AI tool. The adoption rate is accelerating faster than any previous general-purpose technology, including personal computers and the internet.

That sounds like progress. And in some ways it is. But the more useful number is buried further down in the data: the median small business is now using five AI tools. Not one. Five. And most of them are running those tools in parallel, without a clear workflow connecting any of them.

The difference between using AI and getting value from AI

There is a gap between adoption and integration, and most small businesses are sitting in it right now. They have a ChatGPT subscription, a Zapier account, maybe an AI email tool. Each one does something useful in isolation. None of them talk to each other. None of them are connected to the actual workflow that drives revenue.

The JPMorgan data backs this up. In 2019, 89 percent of small businesses using paid AI tools were paying for exactly one. By 2025, that dropped to 72 percent, with 9 percent now paying for three or more. The trend is clear: businesses are accumulating tools. What the data does not show is whether those tools are producing measurable outcomes.

The question is not whether you are using AI. The question is whether the AI you are using is connected to the work that actually moves your business forward.

Why employer firms are pulling ahead

The study found a persistent gap between employer firms and solo operators. By late 2025, 26 percent of employer firms had adopted paid AI tools, compared to 15 percent of nonemployers. And the gap held up even when controlling for revenue.

The researchers concluded that having additional people to implement and integrate AI matters more than having the budget for it. A solo operator managing every function of the business simply does not have the bandwidth to evaluate, deploy, and iterate on AI tools on top of everything else. The bottleneck is not money. It is capacity.

This is the real argument for getting outside help with AI implementation. Not because you cannot figure it out, but because figuring it out while running the business is a different problem than just figuring it out. The businesses pulling ahead are the ones that treated AI implementation as a project with dedicated attention, not a side task.

What this means practically

If you are already using AI tools, the next question is not which tool to add. It is whether the tools you have are connected to a workflow that produces a result you can measure. If you cannot point to a specific outcome, a specific time saved, a specific lead captured or task eliminated, the tool is probably running in the background without doing much.

The businesses that will be ahead in 12 months are not the ones with the most tools. They are the ones that built a small number of workflows that actually work, and then kept improving them. That is a different mindset than tool collecting, and it is the one that produces results.