AI potential does not tell the whole adoption story
A workforce can have tasks suited to AI while the business has yet to put the technology to work.
Save PDF / Print opens your browser’s print options. Choose “Save as PDF” for a copy.
The evidence
The authors link occupational measures of AI exposure to reported firm adoption. Higher exposure is associated with greater adoption, but it accounts for only a modest share of differences among firms. The relationship changes across industries, business sizes, and the exposure measure selected.
- Exposure describes the potential relevance of AI to work; adoption describes reported use.
- Firms with similar occupational exposure can make different adoption decisions.
- Industry and firm characteristics matter when interpreting an exposure measure.
What this means for your business
Economic-development organizations can use exposure to identify questions worth investigating, then examine actual adoption and local needs separately. Entrepreneurs could explore training, implementation, and workflow support where businesses identify a need. Demand must be validated locally; this study does not establish a market size for those services.
What this study cannot establish
The analysis focuses on single-unit firms and should not be treated as representative of all firms. Its relationships are associations, not causal effects. An exposure score is not a local adoption rate or a forecast of job losses.
Are you measuring what AI could do, what businesses already do, or what support they need?