Researchers warn that AI adopters are not representative of the broader economy. Companies adopting AI were already larger, faster growing, more technical, and more likely to receive venture capital backing before implementing the technology, making simple comparisons with non-adopters misleading. To account for this, the study compares early adopters to similar companies that had not yet adopted AI rather than companies that never adopted it.
The report also found that AI adoption remains concentrated in knowledge-intensive industries. Information companies saw the highest adoption rates, followed by finance and professional services, while sectors such as hospitality, arts and healthcare lagged far behind.
Ramp said his research is one of the first to combine observed corporate spending on AI with company-level workforce records, allowing researchers to measure AI adoption based on actual purchases rather than surveys or estimates of occupational exposure. The company defines adoption as three consecutive months of at least $100 in AI vendor spending, and the intensity of adoption is measured by AI spending per employee during the first three months after implementation.
The authors say the results should not be interpreted as evidence that AI triggers hiring, but rather as evidence that companies that make substantial and sustained investments in AI are currently growing faster than comparable companies. They argue that the findings suggest that the initial economic impact of AI may have less to do with replacing workers and more to do with enabling expansion in companies capable of effectively integrating the technology.
