Research

AI adoption in operations statistics, 2026

Operators and lower-middle-market PE-backed companies keep hearing that “everyone is using AI.” The original surveys say something narrower: use is common in large-firm surveys, still thin in nationally representative firm counts, and concentrated in marketing and writing rather than in the workflows that run the back office.

Last updated . Every figure below is the original researcher’s number, linked to the source. Nothing here is estimated, averaged into a new rate, or restated as Pala’s own research except the marked index.

Assessment page

How many U.S. firms used AI in 2026?

The Census Bureau’s Business Trends and Outlook Survey (BTOS) is the nationally representative employer-firm series. Two wordings are in circulation. Through October 2025 the question asked about AI “in producing goods or services.” From 17 November 2025 it asks about AI “in any of its business functions.” Those series are not joined.

18% of U.S. firms used AI in a business function during November 2025–January 2026, or 32% when the same survey is weighted by employment. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

Adoption in that supplement is expected to reach 22% of firms within six months. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

Overall AI usage hovered between 17% and 20% from 14 December 2025 to 3 May 2026, with 19.8% of businesses using AI as of 3 May 2026. Adam Grundy, Cory Breaux, and Dhanapati Khatiwoda, 2026

Between 20% and 23% of businesses expected to be using AI in the next six months over that same December 2025–May 2026 window. Adam Grundy, Cory Breaux, and Dhanapati Khatiwoda, 2026

As of 3 May 2026, AI use was 39.7% in Information and 33.9% in Finance and Insurance, both above the 19.8% national rate, while Retail Trade was around 14%. Adam Grundy, Cory Breaux, and Dhanapati Khatiwoda, 2026

What share of 50–500 person companies use AI?

Pala’s buyers sit in the 50–500 employee band and in PE-backed operating companies of similar scale. Census publishes size cuts around that band. The Federal Reserve’s Small Business Credit Survey covers employer firms with 1–499 employees and uses a broader AI definition, so the two rates are not interchangeable.

Among firms with 50 or more employees, 24.3% reported current AI use on a firm-weighted basis and 36.6% on an employment-weighted basis in the November 2025–January 2026 BTOS supplement. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026, Table C.5

The same table shows 27.3% firm-weighted current use at 100 or more employees and 30.6% at 200 or more employees. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026, Table C.5

Firms with 250 or more employees reported a 31% adoption rate, nearly double the 18% seen in micro-firms with 1–10 employees. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

Firms with 100 to 249 employees had a 25% current use rate and expected future adoption of 34%. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

In the collection period ending 3 May 2026, 32% of firms with 100 to 249 employees and 37% of firms with at least 250 employees reported using AI. Adam Grundy, Cory Breaux, and Dhanapati Khatiwoda, 2026

Less than 20% of firms with four or fewer employees reported using AI in that December 2025–May 2026 window. Adam Grundy, Cory Breaux, and Dhanapati Khatiwoda, 2026

46% of small employer firms (1–499 employees) reported that the business or its employees currently use AI, an additional 15% planned to begin using it in the next 12 months, and one-third had no plans to use AI. Federal Reserve Banks, 2026 Report on Employer Firms

Of those SBCS firms that use AI, about half said they are experimenting, 44% had partially integrated AI into business processes, and 7% had fully integrated it. Federal Reserve Banks, 2026 Report on Employer Firms

Is AI in operations, or still mostly in marketing?

Function-level Census data is the cleanest public view of whether AI has reached the operating system of the firm. The Federal Reserve survey asks about tasks rather than departments. Both point to writing and marketing first.

Among firms already using AI in at least one function, sales and marketing is the most common use, at 52%, followed by strategy and business development at 45%, information technology at 41%, and R&D at 40%. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

57% of those functional AI users integrate AI in three or fewer business functions, and 24% use it in only one. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

37% of functional AI users are “minimalist adopters” with low use probabilities across all functions; only 4% of functional users are “comprehensive adopters” with high use across almost the entire functional spectrum. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

Among small employer firms that use AI, the most common tasks are writing or marketing (83%), individual productivity (61%), and planning or analysis (51%). Federal Reserve Banks, 2026 Report on Employer Firms

70% of potential value from AI is concentrated in core business functions such as sales and marketing, manufacturing, supply chain, and pricing. Jessica Apotheker, Vinciane Beauchene, Nicolas de Bellefonds, and coauthors, 2025

Among firms shifting task structures with AI, 66% engage exclusively in augmentation, and AI-related employment decreases occur in only 2% of firms. Kathryn Bonney, Cory Breaux, Emin Dinlersoz, Lucia Foster, John Haltiwanger, and Aditya Pande, 2026

How many companies have scaled AI past a pilot?

Large-firm executive surveys report much higher “any use” rates than Census. They also show that use is still mostly a pilot.

88% of respondents say their organizations regularly use AI in at least one business function, compared with 78% a year earlier. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

Approximately one-third of those respondents say their companies have begun to scale AI programs; nearly two-thirds have not yet begun scaling AI across the enterprise. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

23% report their organizations are scaling an agentic AI system somewhere in the enterprise, and an additional 39% have begun experimenting with AI agents. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

In any given business function, no more than 10% of respondents say their organizations are scaling AI agents. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

29% of respondents at companies with less than $100 million in revenue have reached the scaling phase, compared with nearly half of those at companies with more than $5 billion in revenue. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

39% of respondents attribute any level of EBIT impact to AI, and most of those say that less than 5% of EBIT is attributable to AI use. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

Only 5% of companies in BCG’s 2025 study of more than 1,250 firms worldwide are achieving AI value at scale, while 60% are not achieving material value at all. Jessica Apotheker, Vinciane Beauchene, Nicolas de Bellefonds, and coauthors, 2025

In BCG’s 2026 Applied AI Index, almost 50% of companies now generate value from AI: 7.5% are “future-built” and another 41% are scaling. Jessica Apotheker, Vinciane Beauchene, Nicolas de Bellefonds, and coauthors, 2026

Does AI raise productivity in operations work?

Measured productivity gains show up in structured work with a clear output. They do not, by themselves, mean the firm has redesigned the workflow.

Access to a generative AI conversational assistant increased customer-support productivity, measured as issues resolved per hour, by 14% on average across 5,179 agents. Erik Brynjolfsson, Danielle Li, and Lindsey R. Raymond, 2023

Novice and low-skilled agents in that same study improved 34%, with minimal impact on experienced and highly skilled workers. Erik Brynjolfsson, Danielle Li, and Lindsey R. Raymond, 2023

71% of small employer firms that use AI said its use led to increased productivity, while the vast majority reported no change in labor costs because of AI. Federal Reserve Banks, 2026 Report on Employer Firms

After accounting for firm and owner characteristics, AI-using small employer firms had a 13 percentage-point higher net probability of expected employment growth than non-users. Will Aarons and Asani Sarkar, 2026

After accounting for firm and owner characteristics and past-twelve-month revenue changes, AI users still had a 14 percentage-point advantage over non-users in net expected revenue growth. Will Aarons and Asani Sarkar, 2026

32% of McKinsey respondents predict an overall workforce reduction of 3% or more in the coming year from AI, 43% expect little or no change, and 13% predict an increase of that magnitude. Alex Singla, Alexander Sukharevsky, Bryce Hall, Lareina Yee, and Michael Chui, 2025

What does PitchBook show for the U.S. middle market?

PitchBook’s public middle-market reports measure deal activity, not AI adoption inside portfolio companies. They still define the operating landscape for lower-middle-market PE: a large stock of held companies, more smaller deals, and value creation that has to come from the business rather than multiple expansion. Pala does not treat these as AI-use rates.

U.S. private equity’s middle market reached $410.7 billion of deal value across an estimated 4,018 transactions in 2025, up 8.5% and 16% year over year. Garrett Hinds, Jinny Choi, and Kyle Walters, 2025

Middle-market exits rose to an estimated 1,022 transactions totaling $140.4 billion in 2025. Garrett Hinds, Jinny Choi, and Kyle Walters, 2025

Middle-market fundraising fell more than 40% year over year to $94.8 billion in 2025, the weakest year since before the pandemic. Garrett Hinds, Jinny Choi, and Kyle Walters, 2025

The 2026 Pala AI Operations Index

This is a Pala original finding. It is a subtraction from a published Census table, not a new survey.

The 2026 Pala AI Operations Index found a 12.3 percentage-point concentration gap among U.S. employer firms with 50 or more employees: 36.6% of employment in that size band sits in AI-using firms, while 24.3% of those firms report using AI.

Method: unweighted subtraction of two published estimates in Bonney, Breaux, Dinlersoz, Foster, Haltiwanger, and Pande, U.S. Census Bureau Center for Economic Studies working paper 26-25R (October 2026), Table C.5, row “50+”: employment-weighted current AI use (36.6%) minus firm-weighted current AI use (24.3%). No interpolation between size bands, no blending with Federal Reserve or McKinsey surveys, and no conversion into a “typical 50–500 person company” rate. The 50+ cut is the closest published Census threshold to Pala’s 50–500 employee operating companies. The gap means AI use in this band is concentrated in the larger employers inside it — which is why an assessment still has to map the specific workflow rather than assume the firm has already adopted.

Source table: CES-WP-26-25R, Table C.5.

What this means for an assessment

National “AI adoption” figures mix a 19.8% Census rate for any business function with a 46% Fed rate that includes stand-alone apps and features of software the firm already uses, and an 88% McKinsey rate from a survey skewed toward larger organizations. None of those numbers tells a COO whether invoice intake, dispatch, or support triage will pay. That is the AI Opportunity Assessment: two weeks, $7,500, credited toward implementation, with a ranked list of workflows rather than a company-level maturity score.

For new owners of acquired SMBs, the same gap shows up as inherited spreadsheets and one-person workflows. Post-acquisition operations starts with that map. For where work usually pays first, see where AI pays off first, finance, and customer operations.

Start with the workflows that would actually pay.

The AI Opportunity Assessment maps which operations, finance, and support paths are worth automating first, and which gaps would break a build.