US Corporate AI Spending Hits Record in October as Growth Slows

What happened

US corporate spending on artificial intelligence reached a new high in October 2026, but the rate of growth slowed to 1.65 percent, according to the Ramp AI Index as reported by The Decoder. Businesses are still committing more money to AI tools, though the increase is modest.

The index also points to a narrow base of demand. Roughly one percent of companies still account for 80 percent of the revenue earned by OpenAI and Anthropic, two of the leading AI model providers.

Customer support is emerging as a new area of growth. The reporting cautions, however, that it may carry lower margins than coding, which has been the main driver of spending so far.

Why it matters

For leaders in experience and digital transformation, the significant question is less whether AI adoption is continuing and more where the value is concentrated. When spending depends on a small group of heavy buyers, both providers and adopters carry concentration risk, and the growth case rests on a few use cases. Customer support is a logical next frontier because it sits directly on the service interaction, where AI can change how enquiries are answered, routed and resolved.

The move from coding to customer-facing work also changes how success is judged. Developer tools are typically assessed on productivity. Support tools must be assessed on resolution, customer satisfaction and cost to serve. Lower margins mean the operating model around a deployment, including staffing, escalation paths and measurement, will matter as much as the model itself.

By the numbers

• 1.65 percent growth in US corporate AI spending in October 2026, according to the Ramp AI Index.
• 80 percent of OpenAI and Anthropic revenue comes from roughly 1 percent of companies.
• October 2026 is the month in which US corporate AI spending reached a new high.

The Renascence take

Most readers will focus on the tension between a record level of spending and a slowdown in growth. The more useful reading is about distribution. A small number of heavy buyers are doing most of the work, and the next wave depends on whether service leaders can turn AI from a productivity tool for developers into a dependable experience layer for customers.

This is really a concentration story, not a spending story. If customer support is going to matter, it will not be because a model looks impressive in a demo; it will be because a resolved query feels effortless to the customer and costs less to deliver. Behavioural economics suggests people judge an interaction in the moment, on whether their problem was solved and how they were treated when it went wrong, not on the technology behind it. Operators should therefore measure AI in support by resolution quality and service recovery, rather than by how much of the workload it touches.