AI Job Displacement Fears Miss What Main Street Did

Small businesses that started using artificial intelligence added staff faster than comparable businesses that did not, according to research Gusto published on September 10. The payroll company tracked 1,593 adopters against 669 firms that knew about AI but had not deployed it. Adopters expanded payroll roughly 7% faster over the following year, which cuts against most ai job displacement forecasts.

The finding matters because it rests on actual payroll records rather than survey sentiment. Gusto followed up with respondents to its own 2025 survey on small business machine learning use, then watched what their pay runs did. For founders deciding whether automation means a smaller team, that is a more useful signal than a headline.

What the Payroll Data Shows

The gap widened as time passed. It measured about 4.5% early on, then reached 6.8% once four quarters had passed. Gusto reads that trajectory as a sustained pattern rather than a one-quarter blip.

Sample size is modest but the method is sound. Comparing adopters against firms that were aware of the same tools controls for a business owner’s general appetite for technology, which is the usual weakness in this kind of study.

The Smallest Teams Gained the Most

Companies under ten employees showed roughly 10% more growth than their peers. Gusto translates that into a nine-person shop adding one extra person inside a year. At that scale, one person is a meaningful change in capacity.

There is a plausible mechanism here. Very small teams already stretch people across tasks outside their job description, so software that absorbs the administrative overflow returns hours to the work someone was hired to do. Owners appear to have spent those recovered hours on serving more customers.

That reading is consistent with recent small business hiring data, which has shown steady if unspectacular job additions through the year.

Clinics and Kitchens, Not Coding Teams

The strongest sector was health care and social assistance. Among practices under ten employees, mental health clinics and dental offices led the way, running nearly 20% above the comparison group on headcount. Their hires were care providers such as hygienists and therapists rather than technical staff.

Accounting firms, manufacturers, and restaurants all posted gains too. Across the full sample, the added roles clustered in hands-on and customer-facing work, from front-desk and admin support to kitchen staff, skilled technicians, and teachers. Almost none of it was AI-specific headcount.

That distribution undercuts the assumption that AI benefits flow mainly to software companies. It also suggests adoption succeeds where a specific bottleneck exists, a point that echoes what happens with AI in the workplace when leaders actually use the tools themselves.

How to Read This Before You Hire

Do not treat the 7% figure as a promise. Treat it as evidence that automation and hiring are not opposites, which frees you to evaluate each on its own merits. The question is which constraint currently limits your revenue.

Name that constraint before you buy anything. If invoicing and scheduling eat twelve hours a week, software is the answer. If you are turning away customers because nobody can serve them, a hire is the answer, and no tool changes that.

Measure the freed hours afterward. Track where the recovered time actually goes for a month, because time that quietly disappears into more admin is not a productivity gain.

The Caveats Worth Naming

Gusto sells payroll software to more than 500,000 small businesses, so it has a commercial interest in a positive small business story. The underlying payroll data is still the strongest available evidence on this question, but the framing deserves the usual skepticism.

Correlation is also not causation here. Businesses confident enough to adopt new software may have been growing for reasons the comparison group did not share, even with the awareness control in place. For a broader read on job creation and destruction rates, the Bureau of Labor Statistics publishes quarterly business employment dynamics figures.

Sentiment remains softer than this data suggests, as the latest small business optimism reading showed. The founder takeaway is narrow and useful: adopting AI has not been shrinking Main Street payrolls, so stop budgeting as if it must.

AI Job Displacement Questions Founders Ask

Does adopting AI mean I will hire fewer people?

Gusto’s payroll data points the other way. Adopters’ headcount grew roughly 7% faster than the comparison group over the year that followed.

Which businesses gained the most?

Health care practices under ten employees, which ran nearly 20% above the comparison group.

What kind of roles did adopters add?

Mostly customer-facing and hands-on work, including care providers, technicians, cooks, and administrative support.