AI spend per employee slumped at top firms in August — summer doldrums or a warning sign?
TechCrunch Tim Fernholz ● Covered by 4 sources
Businesses’ AI spending barely moved in August, per Ramp. That’s a possible vacation lull — or a sign the boom is cooling.
Based on reporting by TechCrunch, Tim Fernholz — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
AI spending at 70,000 companies tracked by Ramp barely budged in August. The payments company says 56% of its customers paid for AI products, up just 0.4% from July. That’s not a collapse. But after the breakneck run-up that’s funded the current AI buildout, even a tiny slowdown is getting attention.
Ramp has seen this movie before. Last year, its AI index showed little or no adoption growth from August through October, then picked up again toward the end of the year. This time, though, the stakes feel larger because the big spending binge on AI infrastructure is built on a simple bet: usage keeps climbing, revenue follows, and the chips and servers eventually earn their keep.
That’s where the latest numbers get uncomfortable. Ramp economist Ara Kharazian says AI spend per employee among the top 1% of firms in his sample fell nearly 10% to $7,205. He also points to lower token prices. OpenAI and Anthropic have cut prices, pushing average token costs down to $0.68 per million tokens from a March peak of $1.15 per million tokens. If customers are paying less per token and not making up the gap with volume, the math gets less friendly for model makers.
There’s a broader wrinkle here too. Ramp’s customers skew technical, so the figures probably overstate adoption compared with the market at large. A U.S. Census Bureau survey updated on August 23 found just 22% of businesses report using AI. And only 6.4% of AI-spending businesses used model-serving or inference platforms in August, even though that share is rising. So the competition between OpenAI and Anthropic may be making AI cheaper and easier to buy, but it’s also pressuring the very spend that frontier labs and hyperscalers were counting on.
The irony is neat and a little brutal. AI is getting more accessible, yet the companies betting hundreds of billions on chips and models are the ones staring at the bill. August could be a vacation blip. It could also be the first sign that the market is getting picky, which is usually where the easy money story starts to sound less easy.
My take — AI-written commentary, not fact-checked reporting
This is the part where the hype crowd discovers that cheaper AI is not the same thing as endless AI spend. If the top firms are already trimming spend per employee while OpenAI and Anthropic slash token prices, the industry’s favourite bedtime story starts to wobble. The real winners may be the businesses using AI, not the ones selling the plumbing to everyone else.
Read more about this at: TechCrunch