Big Tech Margins Hit 25% While the Other 493 Stay Near 10%: Apollo Sees No AI Payoff Yet

News Summary

Apollo Global Management's chief economist, Torsten Slok, says the corporate world outside the technology sector has so far seen no measurable financial payoff from artificial intelligence. In a research note highlighted by TechCrunch at 9:57 AM Pacific Time on October 25, 2026, Slok argued that profit margins and revenue per worker at most large U.S. companies look flat, even as AI spending and enthusiasm keep climbing. His conclusion: the technology "has delivered zero financial payoff to corporate America outside of tech."

What Slok Actually Said

The note, titled "No signs of AI payoff outside the tech sector," focuses on what analysts call the "S&P 493": the S&P 500 index with seven giants removed (Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla). By stripping out the companies that build and sell AI, Slok isolates the firms that are mainly trying to use it. In that group he finds no evidence of better profit margins or higher revenue per worker.

Slok also pushed back on a popular explanation, that non-tech companies are simply a few quarters behind the tech leaders. He doubts that gap will close easily, arguing it will be hard for corporate America to turn cheap machine intelligence into faster revenue growth and fatter margins.

He grounded that view in the limits of the technology itself. "AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness," he wrote, adding that "AI may be a great calculator, but in business, intelligence is mostly judgment." TechCrunch noted that Apollo used to own TechCrunch. The TechCrunch piece did not publish specific percentages, sample sizes or dollar figures from the note.

The Numbers Behind the Argument

Earlier coverage of Slok's analysis supplies the data. Fortune reported on July 6, 2026 (Eastern Time), citing Bloomberg and Macrobond data, that the profit margin of the Magnificent Seven rose from roughly 15% to about 25% between the first quarter of 2023 and the first quarter of 2026. Over the same period, the margin for the rest of the S&P 493 hovered around 10%, and the broader Bloomberg 500 Index held steady at about 12%.

In other words, the profit surge from AI so far sits almost entirely in the companies that supply the technology, chips and cloud capacity.

Why Adoption May Be Slow Outside Tech

Slok's July commentary, as summarized by Fortune and The Decoder, explained why the gap could persist. Software and technology firms can fold AI into their products with relative ease. Much of the rest of the economy faces regulatory hurdles, data-protection rules and the hard work of redesigning workflows. Healthcare, banking, energy, pharmaceuticals and manufacturing were named as sectors where process overhauls and privacy requirements could delay gains "well beyond what the market currently projects."

Measurement is another problem. In knowledge work, productivity gains are hard to quantify, so managers cannot easily act on them. Time saved can be absorbed into daily routines instead of showing up in financial statements. Slok summed it up this way: "The key issue is the length of the ROI runway outside the tech sector."

Supporting Evidence From Other Studies

Fortune's article pointed to several data points that fit the picture:

• A 2025 MIT study found that only 5% of companies saw meaningful returns from generative AI pilot projects.
• A 2026 Boston Consulting Group survey of about 12,000 frontline employees found that 42% reported saving eight hours a week with AI, but most received little guidance on how to use that time.
• In one insurance-claims case at Ricoh, costs were roughly three times higher than manual work, even though division productivity rose threefold.
• Nvidia's Bryan Catanzaro said the cost of AI still far exceeds the cost of human labor.
• Ford hired 350 veteran engineers to train junior staff and improve its AI tools.

These examples show that individual workers can feel faster while the company-wide financial result stays unchanged, at least for now.

Why Investors Should Care

Slok's concern is as much about markets as about technology. Valuations of AI companies depend on the expectation that margins will rise across the broader economy. If the productivity boost takes five years rather than five months, he warns, expected earnings and actual cash flows could diverge sharply, risking what he calls a "painful repricing." He also noted that companies tend to slow AI spending if they do not see a return quickly, and that falling token prices could limit revenue for the large cloud providers.

He does not claim AI will never pay off. His view is that gains may eventually arrive, but possibly after expectations have already been priced in.

A Note on Wording

The headline phrase "zero financial payoff" is the strongest reading of Slok's note. It refers to company-wide profit margins and revenue per worker in the S&P 493, not to every individual project or team. Aggregate margins can stay flat even when particular use cases save time or money, which is why the MIT, BCG and Ricoh examples can all be true at once.

What to Watch Next

Readers following this story can track a few simple indicators: whether the S&P 493 margin moves above its roughly 10% level in coming quarterly reports, whether revenue per worker begins to rise, and whether companies keep raising AI budgets if returns stay invisible. For students of technology, the episode is a useful lesson in the difference between a tool being impressive and a tool changing an organization's economics, which often takes new processes, training and measurement rather than software alone.