AI adoption in business grows steadily but unevenly

Strong investment in the development of large language models have expanded artificial intelligence capabilities in recent years. This rapid evolution and adoption of AI has sparked both deep concern and enthusiasm for this technology. Yet, between these two extremes lies a more nuanced reality.

High-frequency data from the Census Bureau’s Business Trends and Outlook Survey (BTOS) suggest a steady increase in the number of businesses integrating AI into their operations. However, there isn’t a one-size-fits-all adoption process. A special AI supplement published by the Census Bureau in April offers additional details on the characteristics of businesses that are—and aren’t—using AI. Overall, AI use remains largely experimental, and many businesses remain on the sidelines.

The trends

In late 2025, the share of businesses that reported using AI in producing goods or services was about 10 percent, double the rate in mid-2024, according to the BTOS.

Between September 2023 and September 2025, the share of firms planning to adopt AI was slightly higher than those currently using it. Figure 1 shows that firms are likely following through on their intentions and steadily implementing AI.

While some businesses have integrated AI into core production processes, others are experimenting around the edges. A recent modified version of the Census Bureau survey question captures this reality: Since November of last year, the BTOS question now includes AI use in any business function, rather than only in core production.¹ With this broader focus, the share of firms using AI in the first months of 2026 jumped to 20 percent. That is double the rate of those who were using it for core production purposes at the end of 2025.

State-level data indicate that on average, the Fed’s Ninth District is keeping pace with the national trend.² Approximately 18 percent of businesses in this region have integrated AI into their operations. Minnesota leads the way at 20 percent (Figure 2).

Which businesses are using AI?

Implementing new technology often requires financial investment. It also introduces a learning curve that larger businesses may be better equipped to manage. The AI supplement published in April shows that 30 percent of U.S. firms with at least 250 employees said they use AI, while 17 percent of firms with fewer than 20 employees do.

This gap may also reflect differences in how well AI fits individual businesses’ operations. Compared with smaller firms, large firms may handle higher volumes of repetitive work. For example, a firm handling thousands of customer inquiries or invoices may find significant value in AI automation.

AI adoption across sectors varies and likely depends on the extent to which existing workflows are digitized and the abundance of data. Nearly 4 in 10 American businesses in information and professional services sectors reported using AI in their business functions. Firms in educational services as well as the finance and insurance sector trailed very closely behind.

Notably, the manufacturing and health care sectors have experienced significant growth.

Data collected between 2023 and 2025 show that AI use in producing goods or services more than doubled in manufacturing, and more than tripled in health care. In the April supplement, roughly 12 percent of manufacturing businesses and 20 percent of those in health care reported using AI.

Businesses in more labor-intensive sectors are less likely to have adopted AI. Less than 10 percent of businesses in agriculture, transportation, accommodation and food services, and construction reported AI use.

How are businesses using AI?

The BTOS AI supplemental data also show how firms deploy the technology across business functions and its potential effects on employment.

AI dissemination appears to be more task oriented than industry oriented. Businesses are likely adopting AI for efficiency with specific tasks in core production or in supporting functions like marketing, customer service, and human resources.

For example, generative AI has streamlined the production of images, videos, and text. In the April survey supplement, about 15 percent of firms reported using AI for sales and marketing, and nearly 10 percent for public relations and communications (Figure 3).

Unsurprisingly, IT functions also have high AI integration. Digital workflows and access to data are already well established across these tasks, facilitating its implementation.

AI adoption is lower in more specialized functions. For instance, legal and compliance work involves expert judgment and carries significant professional liability, perhaps making businesses more cautious about delegating decisions to AI.

Displacing workers or helping them?

One of the biggest questions surrounding AI is its effect on the labor market, especially whether it will displace workers.

At least in the short term, employment impacts appear to be limited. The survey supplement asked how AI use affected businesses' total employment in the six months prior to survey participation, and roughly 96 percent of companies using AI nationwide reported no change. The remainder were split evenly between those who increased and those who decreased head counts.

Where AI does affect work, it primarily supplements employee tasks rather than replacing them (Figure 4). However, 10 percent of AI-using firms reported that AI is now performing tasks previously done by employees. Given that adoption is still unfolding, these figures deserve ongoing attention.

More broadly, while more businesses expect to use AI in the coming months, a large share remain on the sidelines.

Among those not planning to use it, the most common reason cited was that AI did not apply to their operations (Figure 5). Roughly 20 percent have safety concerns, and a similar share lack knowledge about AI’s capabilities, which highlights a learning curve that may shape adoption patterns.

Overall, there is nuance between the extremes. AI is neither transforming all business overnight nor is it sitting idle. Instead, AI adoption remains experimental, task-specific, and uneven across firms, and its effects on labor and productivity are still unfolding.

Endnotes

1 On October 26, 2023, the Census Bureau Business Trends and Outlook Survey (BTOS) started publishing two core questions to measure the use of AI. Respondents were asked about AI use within the past two weeks and plans to use AI within the next six months. Starting with the November 17, 2025, collection, the BTOS updated its two core questions with broader language to help respondents capture their business use of AI. Specifically, business use of AI in “producing goods or services” was replaced with “any of its business functions.”

2 Ninth District estimates are based on the mean of the published shares for the states of Minnesota, Montana, North Dakota, and South Dakota.