Ramp, a corporate spend management company, found the median monthly AI spend among the top 1% of adopters was $7,401 per employee, compared with $12 for the median adopter. The roughly 620-fold gap appears in its “Summer 2026 Business Spending Report,” which also adds spending-per-employee measures to the company’s AI Index.
The quarterly report analyzes billions of aggregated, anonymized transactions from more than 70,000 businesses using Ramp Bill Pay and corporate cards. AI spending includes large language model subscriptions, coding-agent subscriptions, API tokens and GPU cloud and infrastructure spending.
Measuring adoption versus intensity
The AI Index found that 56% of U.S. businesses in Ramp’s dataset paid for AI models, platforms or tools in July, about 5 percentage points higher than in its previous quarterly report.
The report noted its focus is now shifting from simple adoption toward intensity, with the updated index tracking monthly AI spend per employee alongside company size, sector, state, funding background and type of AI spending.
In the report, the 56% figure compares with a 22% reading from the U.S. Census Bureau’s Business Trends and Outlook Survey. The two measures use different methods. The Census asks businesses whether they used AI in any business function, whereas Ramp uses purchases of AI services made through its platform.
Venture-backed and small firms lead spending
Spending also varies sharply among companies paying for AI. Venture-backed businesses spend a median $76 per employee each month, compared with $7 at private-equity-backed firms.
Small businesses, which Ramp defines as those with 1-24 employees, spend $23 per employee, compared with $3 at businesses with 100 or more employees.
The shift toward usage-based billing
Ramp also found differences in how companies pay for AI features inside traditional software. Seat-only billing remains the most common model among vendors charging separately for AI, but usage-billed products now account for 36% of embedded AI spending, up from 24% a year earlier. Ramp found the median business billed by usage spends about 11 times more annually on embedded AI products than one paying only by seat.
High-intensity adopters see headcount growth
A separate Ramp study, “A New Look at AI’s Impact on Jobs,” which links its spending data with Revelio Labs workforce records for 21,559 U.S. firms, found high-intensity AI adopters grew headcount by 10% over the two years following adoption.
Entry-level headcount rose 12%, whereas low-intensity adopters showed no statistically significant change.
The analysis also carries an important qualification. Ramp said AI adopters were already larger, more engineering-intensive, more likely to be venture backed and faster growing than non-adopters before adoption.
Ramp’s updated AI Index now puts several key metrics side by side: whether a business pays for AI, how much it spends per employee and what kind of AI spending accounts for that budget.