The adoption of AI by industrial sectors

Artificial intelligence (AI) appears destined to impact the entire economy but could affect industries and companies very differently. Some sectors are more susceptible to the application of AI than others. Within sectors, some firms will exhibit greater expertise in adopting AI than others. Different rates of adoption will affect product prices, quality, and delivery speeds, and thus the fortunes of individual firms.

While it is too soon to know how this will play out, some trends are emerging. So far, the earliest users of AI are firms employing highly paid—and therefore presumably high-skill labor; bigger employers; and service sector companies.

Which companies are moving first to adopt AI?

The data presented here suggest that industries employing more highly skilled workers are among the earliest adopters of AI, while those with lower skilled workers are embracing AI more slowly. We see this by taking pay as a proxy for skills. Figure 1 compares average hourly earnings in 2025 with AI adoption by companies in different broad sectors, as reported by the US Census Bureau's recent Business Trends and Outlook Survey (BTOS). The figure shows a modest positive correlation between the two. The early AI users include employers in the fields of information; finance and insurance; and professional, scientific, and technical services. As of late 2025, more than 30 percent of the firms in these industries reported that they were using AI. Conversely, the laggards on AI use generally tend to pay less, including firms in retail, accommodation, and food services (i.e. stores, hotels, and restaurants).

We find a similar correspondence between hourly wages and AI adoption among manufacturing industries, as shown by the trend line in figure 2. Not surprisingly, the computer and electronics industry ranks highest by both measures. The "AI Supplemental Table" in the Census Bureau's Business Trends and Outlook Survey reports that most AI use by manufacturing firms is applied to tasks outside the production of goods, such as sales and marketing.

Larger employers (those with 250 or more workers) tend to be earlier adopters of AI than smaller ones, according to a FEDS Note analyzing the Census business survey.[1] Over the two-year period between 2023 and 2025, firms with 250 or more employees increased their AI adoption from around 4 percent to around 12 percent, whereas firms with 100 employees or less increased adoption from around 3 percent to 8 percent.

These findings are echoed by a survey of AI adoption covering nearly 6,000 CEOs and other senior managers worldwide, detailed in a paper published by the National Bureau of Economic Research. The authors found that "Larger, more productive, and higher-paying firms are more likely to deploy AI technologies."

Service sectors generally lead as AI adopters, and firms with more employees such as Amazon and Meta are more active AI users, according to the Census business survey. Notably, the average AI use in firms with over 250 employees in the information sector was roughly 73 percent in early 2026.[2]

Employment effects?

While much research and many news headlines have focused recently on the possible effects of large language models (LLMs) on the labor market, our PIIE colleague Jed Kolko reported recently that the evidence so far is inconclusive. "Claims about harmful impacts on particular groups of workers are premature," he added.

We speculate, however, that to the extent that AI adoption improves a firm's performance, faster adopters are likely to gain competitive advantages over their slower rivals. Thus, faster AI adoption could increase the concentration of US employment in large firms by lowering their costs and improving their product characteristics relative to small firms. In 2000, firms with 1,000 or more employees accounted for 37.6 percent of US employment. By 2025, this share had increased to 42.3 percent. Further concentration seems likely from future AI adoption. Indeed, high AI adopters will likely enjoy competitive advantages across multiple dimensions, such as price, speed, and quality, and in turn capture greater market share. As one example, the law firm Freshfields will partner with Anthropic to build legal AI tools that can be sold to other law firms, hoping to thereby expand Freshfield's share of the legal market.

One factor likely to affect the market share impact of AI adoption is the price elasticity of demand, expressed as the percentage increase in quantity of the product demanded for each 1 percent decrease in price (or its equivalent in quality or delivery speed).

Within a sector, products with higher price elasticities experience larger demand gains following a reduction in price. So, we can speculate that high AI adopters in these industries will benefit most because the productivity gains from AI will enable them to cut prices and gain more market share from slower AI adopters.

However, the survey data published by NBER show somewhat higher AI adoption rates for sellers of products with low elasticities of demand.[3]

Many media stories prominently feature layoffs by firms that are early adopters of AI, such as Amazon, Oracle, and Target.[4] Our findings suggest that this emphasis may miss the larger story. Capture of market share by early adopters may on average squeeze laggard firms making them the bigger job losers.

Notes

1. It is worth noting that large firms tend to have more business functions, both in number of staff and diversity of roles, so they are naturally more likely to report "Yes" to the AI adoption question.

2. The estimate based on simple average of surveys (202525–202604) for Sector 51 (Information) and Employment Size Class G (250+ employees), using data from BTOS Sector by employment size class data.

3. This generalization is based on the elasticity estimates of Ahmad and Riker 2019, Ahmad and Schreiber 2024, and AI adoption data in Business Trends and Outlook Survey.

4. A Wall Street Journal story illustrates such reports: "Tens of Thousands of White-Collar Jobs Are Disappearing as AI Starts to Bite."