Top 1% Companies' AI Spending Reaches $7,400 Per Employee
According to Ramp data, AI budgets among U.S. companies show extreme polarization, with the median AI spending per employee in the top 1% of companies reaching a record $7,400 in July.
The top 10% of companies spend about $650 per employee, while the median company spends only about $11.95, a difference of over 600 times.
In recent months, per capita AI spending has more than tripled for the top 1%, top 10%, and median companies.
At the beginning of 2024, the top 1% of companies still had per capita spending below $1,000, indicating a concentration of spending.
Among suppliers, Anthropic accounts for 43.5% of paid U.S. enterprise spending, OpenAI for 39.7%, and SpaceXAI has risen to 4%.
From a market mechanism perspective, AI spending is highly concentrated among a few "whale" companies, with funds flowing towards cutting-edge models and multi-model routing. Buyers are primarily high-intensity adopters, with a structural increase in demand for event-driven computing power and APIs, while median companies lag behind, benefiting high-spending companies.
Supplementary data shows that spending growth is accompanied by some companies shifting to cheaper models to control costs.
Source: Public Information
ABAB AI Insight
Ramp tracks AI spending of over 70,000 U.S. companies through its corporate card and bill payment data, having previously released indices revealing adoption gaps. Historical behavior shows that it considers AI spending a core indicator of corporate digital maturity.
On the capital path, top companies invest thousands of dollars per employee into multi-model and agent systems, motivated by the desire to seize efficiency and innovation advantages. Strategically, they form data and process barriers through high-intensity usage while testing cost optimization combinations.
Similar cases can be seen in the early cloud computing era where "whale" customers dominated spending. Current AI adoption is in a phase of differentiation from experimental subscriptions to production-level high-intensity consumption.
The structural judgment indicates capital concentration: because a few companies bear the vast majority of spending, the mechanism makes model providers' revenue highly dependent on top clients, forcing suppliers to prioritize high-usage scenarios, shifting pricing power from widespread adoption to deep integration and customization.
ABAB News · Cognitive Law
- AI spending follows an extreme power law distribution.
- Whale customers determine demand structure.
- Spending over a thousand dollars per employee is considered true entry.