AI delivers 'zero financial payoff' to most companies, Apollo economist says
On October 24, 2026, Torsten Slok, partner and chief economist at Apollo Global Management, published a note arguing that AI has produced no financial return for companies outside the technology sector. He built the argument around the S&P 493, the S&P 500 without Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. Across that group, profit margins and revenue per worker were flat, while the seven megacaps show both rising. He also found no sign that AI is raising productivity growth. TechCrunch, which reported the note on October 25, also notes that Apollo once owned the outlet.
The note contains five charts and was published as earnings season opened, when investors were watching AI spending closely. Slok wrote that mainstream companies are starting to treat token spending as a cost that weighs on margins rather than a source of growth. He said the AI story will be decided in the S&P 493, and that macroeconomic effects stay limited while agentic AI and token use remain confined to a small part of the economy. Slok's zero reading covers aggregate trends. The note does not separate companies that adopted AI deeply from those that bought licenses and changed little, and it excludes private companies.
Why it matters
For founders selling AI products to mainstream companies, the bar is shifting from usage to measurable profit impact. Slok frames token spending as overhead that buyers scrutinize, so vendors that cannot show revenue or margin gains may face tighter budgets and harder renewals. Futurism reported that the cost of frontier models has climbed enough to push some companies toward cheaper alternatives, which puts pressure on premium pricing.
For investors, the note adds weight to concerns about AI infrastructure spending. Gartner has estimated that over $3.5 trillion in AI infrastructure outlays could need writedowns by 2030 if revenue does not catch up. Diligence should test whether a company's customers can show returns, not only whether they use models heavily. Operators should track return per customer, since cheaper intelligence only helps margins if it converts into higher revenue.
Key points
- 1Slok's note covers the S&P 493, the S&P 500 excluding seven megacap tech companies.
- 2Revenue per worker and profit margins are rising for the Magnificent Seven but flat for the S&P 493.
- 3Slok says mainstream companies increasingly treat token spending as a margin drag rather than a growth driver.
- 4Gartner estimates over $3.5 trillion in AI infrastructure outlays could need writedowns by 2030 without revenue support.
- 5
