Ramp Data Shows AI Spending Stalling Just as Anthropic Nears Its IPO

Companies are using more artificial intelligence than ever, but their money spent on it has flatlined. A new report shows that usage volumes rose while spending stalled in the late summer.

Data from Ramp shows that spending on tokens grew by only 1.5% across August and September. That was a sharp drop from July, when spending grew by 42% from the previous month. The data suggests that investors might be wrong to think artificial intelligence companies will keep growing at breakneck speed as Anthropic prepares for its initial public offering.

Ara Kharazian, Ramp’s lead economist, said almost all spending on the technology comes from venture-backed, high-growth technology companies. He noted that this trend works fine as long as spending keeps going up, but if the market slows down, those firms will slow down at the same time and cause a chain reaction. Kharazian pointed out that the top 1% of firms, which spend the most, saw their spending drop slightly in August and September below their summer peak of $10,000 per employee. That small slowdown among top spenders was enough to slow overall revenue growth even though usage volumes kept rising.

Median firms spend $12 to $15 per employee each month, and the top 10% spend $600 to $1,000. Kharazian doubts everyday businesses like doctor offices or car dealerships will ever reach the $10,000 monthly per-employee mark. October data did show a rebound with token spending up 8.9% from September 23 to October 23, but strong headwinds on growth remain.

Kharazian explained that the slowdown is not caused by a lack of demand, since artificial intelligence use volumes continue to climb. Instead, fierce market competition has made the technology cheaper to use. OpenAI and Anthropic have both cut application programming interface prices and offered enterprise discounts to lock in customers, which has ultimately hurt their revenue targets.

Other major tech firms are also introducing cheap or open-source models, while smaller tools like Jev let users route tasks to lower-cost options. At the same time, companies have grown more skilled at controlling costs by setting employee budgets and using routers to direct tasks to the cheapest models.

These shifting market dynamics have affected major industry players. Bloomberg reported that Anthropic's potential valuation sank from an early October estimate of $2 trillion down to $1.2 to $1.3 trillion, and the company is expected to receive a high junk credit rating. OpenAI also lowered its proposed valuation to $1.25 trillion and cut its fundraising goals by about $10 billion.

Investors hope that future innovations like personal artificial intelligence agents might drive the next wave of hypergrowth, but Kharazian remains skeptical. He noted that personal agents are not as economically productive as coding agents, leaving the industry with slowing spending based on currently available products.

Summary

Ramp’s data reveals that fierce competition and thriftier corporate customers are slowing spending growth even as usage increases. This unexpected shift challenges traditional growth assumptions just as major artificial intelligence companies face lowered valuations and upcoming public offerings.