Apollo Global Management chief economist Torsten Slok has concluded that artificial intelligence has lifted revenue and margins only for the Magnificent 7, while the rest of corporate America has seen no financial benefit. Companies across the rest of the S&P 500, he found, have flat revenue per worker and flat profit margins, with no increase in productivity.

Slok laid out the findings in a note titled “No Signs of AI Payoff Outside the Tech Sector,” published Saturday in Apollo’s Daily Spark.

Top 10% of firms account for 99.5% of inference spending

Slok splits the index in two. The Magnificent 7 are Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla; the S&P 493 is everything else in the S&P 500.

Inflation-adjusted revenue per employee among the Magnificent 7 has climbed to about $780,000. The S&P 493 sits at roughly $525,000 and has not budged. Margins divide the same way: the seven giants run close to 30%, against about 10% for the other 493 firms.

“Despite massive investments and widespread hype around artificial intelligence, the technology has delivered zero financial payoff to corporate America outside of tech,” Slok wrote.

His total factor productivity chart uses the utilization-adjusted series from the San Francisco Fed — an estimate of output growth that added labor and capital cannot explain, adjusted for how intensively resources are used — which shows no acceleration despite the flood of AI money.

The concentration is just as stark on the spending side. In Slok’s data, the top decile of firms accounts for 99.5% of model serving and inference expenses — the computing that runs trained models and returns their responses — and 99% of neocloud expenses, the shorthand for AI-focused cloud providers. The same 10% represents more than 95% of money tied to OpenAI and Anthropic, and 95.7% of cloud computing.

Venture money rules the biggest users: 80.4% of the top 1% of AI spenders are backed by venture capital, compared with 13.1% of the median spender.

Mainstream companies are now treating token spending — the billable units of text that AI models process — like a cost center that is harming their margins and are pulling back, Slok wrote. Corporate America does not need the latest model and will not pay for it, he said. “As as agentic AI and token spending are confined to a small part of the economy, the macroeconomic effects of AI will be very limited,” he wrote.

The S&P 493 will decide the story, the economist concludes

Slok’s spending charts draw on the Ramp AI Index and Ramp Chief Economist Ara Kharazian. Cryptopolitan reported this month that corporate AI spending has dropped since July even as usage jumped about 50% to a record, as price competition between OpenAI and Anthropic made models cheaper.

Slok does not think the laggards simply need more time. Models keep getting cheaper and more commoditized, he noted, and he questions whether ordinary companies can use them to drive faster revenue growth or wider margins.

“AI may be a great calculator, but in business, intelligence is mostly judgment,” he wrote. Judgment is contextual, Slok said, and the answer depends on which human or model gets asked. He added that AI “does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness,” elements he regards as drivers of reputational risk and, ultimately, of revenue and profit margins.

“The AI story will be decided in the S&P 493, not the Mag 7,” Slok concluded. He has found no evidence yet of better revenue growth or margins outside of tech. Future updates to the two datasets behind those charts — the San Francisco Fed’s productivity series and the Ramp AI Index — will show whether the S&P 493’s flat revenue-per-worker and margin lines start to move.

By April, AI-related stocks had reached about 45% of the S&P 500’s market value. Over the past 12 months, the Global X Artificial Intelligence & Technology ETF (AIQ) has risen about 20%, while the iShares U.S. Technology ETF (IYW) is up about 23%.