Apollo Economist Torsten Slok Says AI Has Delivered Zero Financial Payoff Outside Big Tech

Apollo Global Management chief economist Torsten Slok wrote in a note dated October 24, 2026 that artificial intelligence "has delivered zero financial payoff to corporate America outside of tech." Revenue per worker and profit margins are rising for the Magnificent 7 but are flat for the rest of the S&P 500, he wrote in the note.

The largest tech companies keep raising their AI budgets: Meta now expects full-year capital spending of $135 billion to $145 billion and reported negative free cash flow of $2.12 billion for the third quarter, CNBC reported on October 22, 2026. Slok's point is that the money moving through the AI business has so far shown up in tech companies' results, not in those of their customers.

What did Apollo's Torsten Slok say?

Torsten Slok wrote on October 24, 2026 that AI has had no impact on the bottom line of mainstream US companies, and that they are starting to treat token spending as a cost center that weighs on margins rather than as a source of growth. Tokens are the small units of text that AI providers meter, and many bill by them, so costs climb as employees and agents send more requests (Tom's Hardware).

The note makes three claims, each tied to a chart. Revenue per worker is rising for the Magnificent 7 but flat for what Slok calls the S&P 493; profit margins follow the same split; and there are "still no signs of AI boosting productivity growth," he wrote.

Venture backed companies are spending more and more on tokens, but mainstream corporate America "does not need the latest model and is unwilling to pay for it," he wrote. As long as agentic AI and token spending stay confined to a small part of the economy, the macroeconomic effects of AI will be very limited, according to the note.

What do Apollo's charts show?

Apollo's charts, built on Bloomberg and Macrobond data, show the Magnificent 7 pulling away from the rest of the S&P 500 on both revenue per employee and profit margin since 2023, while the S&P 493 lines barely move. These three charts carry no data labels, so the figures below are this site's readings of the lines.

On revenue per employee, inflation adjusted to fourth quarter 2019 dollars, the Magnificent 7 line rises from about $565,000 in early 2023 to roughly $780,000 in the latest quarter shown. The S&P 493 line stays between about $500,000 and $530,000 over the same stretch. On profit margin, the Magnificent 7 reach close to 30% in the latest quarter, while the S&P 493 sit near 10%, in a band of roughly 9% to 10.5% since 2023.

The productivity chart uses total factor productivity (TFP), a measure of output growth that more labor and capital do not explain, taken from the San Francisco Fed's utilization adjusted series, next to output per hour from the Bureau of Labor Statistics. At the end of the chart, TFP growth dips just below zero year over year, and output per hour runs at roughly 2%.

Slok has made a version of this argument before. In comments Fortune reported on July 6, 2026, Magnificent 7 margins rose from about 15% to about 25% between the first quarter of 2023 and the first quarter of 2026, against roughly 10% for the S&P 493. On the new chart, the Magnificent 7 line has climbed further since then.

Who is actually spending on AI?

Most AI spending comes from a small group of companies, according to Ramp data in Apollo's note: the top 10% of firms account for 95.6% of spending on OpenAI and Anthropic and 99.5% of spending on model serving and inference. Ramp is a corporate card and spend management company, and the chart credits its chief economist, Ara Kharazian.

A neocloud is a cloud provider that mainly rents out GPUs for AI work.

A fifth chart splits the heaviest spenders by funding. Among the top 1% of AI spenders, 80.4% are venture backed; among median spenders, 13.1% are, according to the Ramp AI Index data Apollo cites. Apollo titles the chart "Circularity in AI shows up in token spending too."

Adoption, by contrast, is broad. Ramp's AI Index on the St. Louis Fed's FRED database shows that 55.86% of businesses in its sample paid for an AI product or service in September 2026. Read together with Apollo's table, more than half of these businesses pay for AI, yet the bottom 90% of firms account for less than 5% of spending on OpenAI and Anthropic, by this site's reading of both sources. Apollo's chart does not say which period its table covers, so the two figures may not line up exactly in time.

Why does Slok doubt cheaper AI will close the gap?

Slok wrote that the question looking ahead is whether mainstream companies can turn "increasingly cheap intelligence" into higher revenue growth and higher profit margins as models become commoditized, and that this "may be harder than it sounds."

Judgment depends on context, Slok wrote, and AI does not understand accountability, trust, culture, unwritten rules, ethical trade offs or fairness, which shape reputational risk and in the end drive revenue and margins. His conclusion: the AI story "will be decided in the S&P 493, not the Mag 7."

What we don't know yet

• Which period the Ramp spending table covers and how Ramp defines the top 10% of firms; the chart does not say.
• How much of the Magnificent 7's margin gain comes from AI itself rather than their other businesses; the note does not break it down.
• Whether smaller private companies outside the S&P 500 show the same flat results.
• How companies in the S&P 493 and other economists respond to the note.