US companies spent more on AI in October 2026 than in any earlier month tracked by the Ramp AI Index. The record came with a modest 1.65 percent month-over-month increase, and the figures point to a market where spending is concentrated in a small number of businesses and uses.
A record shaped by a wide spending gap
The typical company in Ramp’s data spent $16.32 per employee per month on AI, up by less than a dollar from the previous month. Among the top one percent of companies by AI spending, the median was $10,200 per employee per month, a nine percent increase.
That contrast helps explain why a new high for total spending does not necessarily mean AI is spreading evenly across US businesses. Ramp’s figures reflect payments made by its business customers, and the company says its customer base tends to be larger, more growth-oriented, and more tech-savvy than US businesses overall.
The index draws on payment data from more than 85,000 US companies. It may miss free AI tools and use through personal accounts, while separate token-volume and spending measures cover only companies using Ramp Token Spend Management. Ramp cautions that these data are not necessarily representative, and that adoption may be overstated.
A narrow base drives provider spending
Spending on OpenAI among business customers rose 2.1 percent month over month, while spending on Anthropic increased 0.5 percent. Yet both providers remain reliant on a small top tier: according to Ramp, one percent of companies drive 80 percent of revenue for OpenAI and Anthropic.
Those heavy users are mostly venture-backed companies in technology, finance, and technical services. The source describes adoption outside those fields as barely spreading, a sign that current business demand is not yet broadly distributed across sectors.
Ramp’s adoption figures offer another view of the competitive picture. Anthropic reached 44.8 percent adoption, up 1.1 percentage points, while OpenAI held at 40.6 percent. Overall adoption of AI services on the platform rose to 56.8 percent. These figures describe Ramp’s customers, rather than the full US market.
Software development still dominates, but support is rising
Software development teams accounted for 72 percent of tracked AI spending in October. Their spending had declined from summer highs, but coding remains the biggest identifiable use in the dataset.
Customer support and customer success gained ground. Their share of AI spending climbed from 0.8 percent in June to 13.3 percent in October. The report suggests AI chatbots in customer service may be contributing to that increase, though the source does not establish a single cause.
The shift matters for how companies use AI and how providers earn from it. Support applications tend to favor cheaper, faster models, according to the report, and their growth has not matched the earlier surge in development-team spending. Leadership teams also spent slightly more.
Lower prices and changing economics
The average price per million processed text tokens fell 21 percent last month, according to Ramp. Usage volume is growing faster than spending, so companies can process more tokens without a matching rise in their bills. Ramp attributes the falling prices to competition between OpenAI and Anthropic, as well as open-weight models from China.
For AI providers, an expanding customer base in support could bring lower margins if these workloads rely more on less expensive models. At the same time, the share of spending on the most expensive models—Opus, Fable, Sol, and Astra—edged up to 69.5 percent of total spending in October.
October’s record therefore combines several trends: more AI use, lower average token prices, and spending concentrated among leading companies and software teams. The rising contribution from customer support suggests that the mix of business applications may be broadening, even as the largest share of tracked spending remains tied to development.