How AI adoption by small businesses could reshape the economy – and what that means for investors

AI has largely been framed as a mega-cap story: bigger balance sheets, bigger budgets for research and development, and bigger earnings impact. But the more consequential question for the broader economy is what happens when AI shifts from novelty to operating model change. That matters in today’s macroeconomic environment, where productivity, labor constraints and wage pressures can affect companies differently depending on their size.

That why it’s worth looking past the biggest names in tech and toward the millions of small businesses that collectively shape hiring, service delivery and local competition. If AI becomes a true general-purpose technology across this base – rather than just a handful of pioneering companies – the effects could show up in places investors actually care about, such as worker output, profit margins, pricing power and competitive advantages.

To move beyond anecdotes, our JPMorganChase Institute analyzed proprietary Chase Business Banking data to track how AI adoption by small businesses is evolving today. The recent report examines AI adoption by business age, whether a firm has employees, and – most importantly – whether usage looks more like experimentation or real integration.

For investors, the key takeaway is to understand that wider AI diffusion could broaden to the point where productivity gains show up beyond Big Tech.

Early adoption and rapid acceleration: Newer firms are using AI sooner

One reason that AI adoption by small businesses is worth watching is speed. Indeed, the JPMorganChase Institute analysis mentioned earlier shows small business adoption has increased significantly – from about 5% when generative AI became more widely available in late 2022 to nearly 18% by the end of 2025. That kind of implementation in just a few years suggests that AI isn’t spreading like a niche tool; rather, it’s starting to look like a foundational capability.

Looking under the hood to determine when a business started operating makes the acceleration trend even clearer. According to JPMorganChase Institute data, only 1.2% of small businesses formed in 2019 adopted AI from the outset. By 2025, those “day-one” adoption rates had jumped to 6.5%. That means small businesses not only are adopting AI earlier but are doing so more rapidly. Indeed, the same Institute data reveals that the 2025 cohort reached a 10% adoption threshold 10 times faster than the 2019 cohort.

There are several reasons why newer firms might be integrating AI earlier in their life cycle. Let’s consider them.

The speed of this trend matters. Faster adoption among newer small businesses can compound over time. If more companies are building with AI in their stack from the outset, diffusion can broaden beyond the largest companies. The next question will be whether that adoption is mostly experimentation – or if it is legitimately transitioning to integration that can bolster productivity.

Adoption vs. integration: The difference that drives economic impact

Rising AI adoption is an important signal, but it doesn’t automatically translate into productivity gains. There’s a sizable difference between a business that uses an AI tool for a few months and a business that rebuilds entire workflows around it – from training employees to putting guardrails in place.

The JPMorganChase Institute analysis looks beyond whether a business uses AI at all and instead measures the depth of usage: whether a business pays for AI consistently over time, and whether it pays for multiple types of AI services. Together, these patterns can help identify sporadic experimentation versus consistent, diversified use that is more likely to reflect real operational integration.

Timing also matters: During the rapid adoption wave of 2022–23, the mix skewed more toward experimentation as many businesses tried AI for the first time. However, data shows signs that integration had become more widespread by 2024–25. Indeed, consistent usage has risen relative to sporadic engagement as more firms move from “testing” to “using” AI for routine operations.

Broad productivity gains require integration, not trial subscriptions. Adoption alone doesn’t change the economy, but integration can – especially if it helps a large number of small businesses produce more per hour, improve customer responsiveness or generally compete more effectively.

Individually, small businesses don’t impact gross domestic product. Collectively, though, they make up a large share of firms and employment, so even modest efficiency gains spread across millions of businesses can add up to meaningful shifts in productivity, labor demand and competitive dynamics over time.

Our strategists view this shift toward more consistent, diversified AI usage as a constructive signal for the broader U.S. economy. We expect AI-driven productivity gains to build over the medium term, primarily as firms move beyond experimentation and embed AI into everyday workflows. Our findings suggest this transition is already underway.

Which small businesses find AI practical?

If generative AI is as accessible and low cost as many headlines suggest, it would be easy to assume the earliest and most enthusiastic adopters would be solo operators, or businesses without employees that can use AI to “do more with less.” But the JPMorganChase Institute data suggests the opposite: Small businesses with employees have been more likely to adopt AI than those without.

Specifically, over a quarter of employer firms have adopted AI, compared with about 15% of nonemployer firms. That gap persists even when comparing firms by revenue – which points to something other than affordability as the main constraint to widespread adoption. In practice, the limiting factors may instead be time and implementation bandwidth: Integrating AI into workflows typically requires upfront setup, procedural changes and ongoing oversight. A business with only a few employees can devote some capacity to building out and monitoring AI tools, while a sole proprietor might not have that capacity.

If AI helps businesses respond to customers faster, generate leads more efficiently or streamline back-office work, those advantages can often compound – especially for firms that can systematize usage rather than rely on one-off prompts. Over time, that can increase dispersion among small businesses: Operators who integrate AI effectively may pull ahead on productivity and service levels, while those without the bandwidth to implement may struggle to keep pace.

This split between employer and nonemployer firms also helps explain where demand could increase next. If the bottleneck is implementation – not just access – then the opportunity isn’t limited to AI models themselves. It extends to the “picks-and-shovels” enablers that make AI usable in day-to-day operations – that is, workflow tools, managed services, training and support, governance and quality control, and products that plug AI into existing systems such as payments or bookkeeping.

None of this is guaranteed. AI can help small businesses scale output and improve responsiveness, but the risks are real – especially for smaller firms with limited redundancy. Specifically, poorly implemented AI can introduce errors, compliance or privacy issues, and even reputational damage. The takeaway is that practical adoption depends less on curiosity and more on capacity – the ability to integrate AI into how the business actually runs.

How could AI adoption by small businesses reshape the economy?

If AI moves from occasional experimentation to real workflow integration across small businesses, it could change how firms produce, hire and compete. And because small businesses collectively represent a large base of day-to-day economic activity, small efficiency gains – repeated widely – can compound into macro effects over time.

In practical terms, the impact would likely show up in three places: higher productivity (more output per hour as routine work gets compressed), shifting competition and margins (faster/cheaper service delivery raising the bar and squeezing laggards), and changes in labor demand (AI augmenting some roles while reducing hours needed for repetitive tasks). Over time, if those productivity gains spread broadly across services, they could also influence pricing dynamics, though that’s downstream and far from guaranteed.

One important note for investors: Economy-wide impact depends on sustained integration. If adoption stalls at trials – or if implementation creates errors, privacy/compliance issues or reputational damage – gains may be limited and competitive churn could rise.

The bottom line: What more AI adoption among small businesses means for investors

The investable question is whether AI becomes more deeply embedded across everyday small business workflows – broadening the productivity narrative beyond Big Tech – while also creating disruption risk in parts of the services economy.

In that environment, investors may see:

There are two signals investors may want to watch for: whether usage is becoming truly integrated (consistent and diversified, not sporadic) and whether competitive churn is rising in the most AI-exposed service categories.

If the shift toward integrated AI usage continues, it could support a broader productivity backdrop over time, which is one reason investors may want to monitor diffusion beyond Big Tech.

Read the full JPMorganChase Institute report to dive deeper into the data behind these trends.

View Disclosures