Despite hundreds of billions of dollars poured into datacenters, enterprise licensing, and generative AI pilot programs, artificial intelligence has delivered zero measurable financial payoff to corporate America outside the technology sector. That is the conclusion of an empirical research report published on October 24, 2026, by Dr. Torsten Sløk, Chief Economist and Partner at alternative asset management giant Apollo Global Management.
Analyzing financial filings across the S&P 500, Sløk observed a stark divergence between the 'Magnificent 7' tech conglomerates and the remaining 493 companies that represent the backbone of the American economy. While revenue per worker and operating profit margins have climbed steadily for cloud and hardware providers selling artificial intelligence tools, both metrics have remained completely flat for non-tech firms when adjusted for inflation against 2019 baselines. Furthermore, the Federal Reserve Bank of San Francisco's utilization-adjusted Total Factor Productivity (TFP) series shows no statistical inflection point, indicating that generative assistance has not translated into broader macroeconomic efficiency.
The research highlights a fundamental shift in how corporate chief financial officers view generative AI procurement. According to transaction telemetry from the Ramp AI Index cited in the study, while venture-backed software startups continue accelerating expenditures on API tokens, traditional corporations in retail, manufacturing, logistics, and financial services are freezing deployments. Facing escalating subscription invoices without commensurate revenue uplift, executive committees are reclassifying token consumption from strategic research investments into operating cost centers that actively compress operating margins.
Sløk attributes this commercial friction to the difference between mechanical computation and business judgment. While neural networks excel at rote summarization and code completion, commercial profitability depends heavily on contextual risk assessment, institutional trust, and unwritten cultural norms. Because language models lack accountability and struggle with ethical trade-offs, corporations incur substantial verification and oversight overhead that cancels out anticipated labor savings. Sløk concluded that the commercial destiny of artificial intelligence will ultimately be determined by adoption across the S&P 493, where enterprise buyers are demonstrating little willingness to pay premium prices for commoditized models.