No signs of AI payoff outside the tech sector

Torsten Slok, Apollo's chief economist, published the note on October 24, 2026. Business spending on AI tokens grew just 0.2% in September, down from 42% in July, based on data from Ramp. Some firms are cutting AI spending as token costs are increasingly viewed as a drag on margins. The note also reports that 80.4% of the top 1% of AI spenders in the data Slok cites are VC-backed. The wider data points the same way. A St. Louis Fed review of about 490,000 earnings calls found that executives mostly describe AI productivity gains as something still to come. San Francisco Fed data shows utilization-adjusted total factor productivity grew just 0.07% over the four quarters ending in the first quarter of 2026. An NBER survey of executives found perceived productivity gains are larger than measured productivity gains, likely reflecting a delay in revenue realizations. For a team building an AI company, the test is whether a buyer outside tech sees a measurable margin gain, not whether the model is capable. Products that consume many tokens per customer may face cost pushback, while tools that remove a visible expense are easier to justify. The counterpoint, from economist Brad DeLong, is that benefits to users, new businesses and the possibility of established platforms locking in customers could matter more than measured GDP or non-tech profits.