Topline
About 69% of S&P 500 companies actively deploy artificial intelligence, according to an analysis by Apollo Global Management chief economist Torsten Slok, but he said only about 2% of these companies actively track its impact.
Torsten Slok, chief economist at Apollo Global Management, said many S&P 500 companies have adopted AI, but argued few have proved the return on investment. (Photo by Silas Stein/picture alliance via Getty Images)
dpa/picture alliance via Getty Images
Key Facts
In a blog post for Apollo Global Management on Tuesday, Slok said 69% of S&P 500 companies actively deploy AI technology based on an analysis of second-quarter 2026 earnings, up from 64% for the first quarter.
But Slok argued many companies do not quantify the impact of AI, saying just 29% of S&P 500 companies report a quantified result and only 2% report a metric tracked over time.
None of the S&P 500 companies report AI as its own key performance indicator or profit-and-loss line, Slok said.
Slok cited data by The AI Value Gap, which found at the end of the second quarter of 2026, only about 30% of large companies “can point to a single realised, quantified AI result,” noting “AI value is still mostly claimed rather than proven.”
“The question is no longer who is deploying AI. It is who can prove the [return on investment],” Slok wrote.
what to watch for
Whether the adoption of AI or reporting of AI-related metrics increases following the reporting of third-quarter earnings. Most of the metrics Slok tracked between the first and second quarters rose, including the number of S&P 500 companies that have a stated AI plan, which rose from 68% to 74%. The number of S&P 500 companies that report a quantified AI result rose from 26% to 29%, while the number of companies that track a metric over time rose from 1% to 2%.
key background
Slok has previously argued large companies are spending a lot of money on AI with minimal results to show for it. In an August blog post, he said big companies are “spending heavily on AI. But it is not showing up in profit margins,” noting most of the companies outside of the tech industry are “flat, cyclical or worse.” Slok argued the longer it takes for big companies to generate a return on investment in AI, the “bigger the downside risks to an economy and a market this concentrated in the AI trade.” In a June blog post, Slok said there are “no signs of profit margins rising outside the tech sector” from AI investment, suggesting this has inflated the values of AI companies, which rely on future hopes that return on investment among S&P 500 companies will climb. He suggested companies that do not see a return on investment will eventually slow AI spending, which he called an “early warning that AI implementation could be a bumpier, slower road than expected.”
surprising fact
A study published by Massachusetts Institute of Technology researchers last year found only 5% of AI pilot programs launched at large companies delivered “rapid revenue acceleration,” while 95% stalled and delivered little to no impact on a company’s profits. Aditya Challapally, the lead author of the study, told Fortune many of the AI project failures stem from a “learning gap” among companies, saying organizations struggle to integrate AI tools because they “don’t learn from or adapt to workflows.”
further reading
MIT report: 95% of generative AI pilots at companies are failing (Fortune)
Leave a comment