Mainstream US companies see zero financial payoff from AI, Apollo analysis finds
TechCrunch|AIDEN Editorial|Oct 25, 2026
An analysis by Apollo Global Management revealed that artificial intelligence has yielded virtually zero financial gains for mainstream US companies outside the leading technology giants. While the Magnificent Seven saw significant revenue and margin growth, the remaining 493 S&P 500 companies experienced stagnating productivity and flat profit margins since 2022.
Artificial intelligence investments have delivered virtually zero financial payoff to mainstream corporate America outside the major technology giants, according to an analysis by Apollo Global Management reported by TechCrunch on October 24, 2026. While infrastructure suppliers and hyperscalers pour billions of dollars into data centers and hardware, enterprises adopting these tools are accumulating expenses without achieving measurable gains in productivity. An examination of the S&P 500 showed that while the Magnificent Seven tech firms—Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla—expanded their revenue and profits rapidly, revenue per employee and operating margins across the remaining 493 companies have remained flat since 2022.
The divergence between leading technology providers and broader industry participants appears stark across financial and macroeconomic metrics. Inflation-adjusted annual revenue per employee for the 493 non-tech companies hovered between $525,000 and $550,000, extending a seven-year stagnation, while the top seven tech firms increased their figure from roughly $550,000 to between $780,000 and $800,000. In addition, operating margins for mainstream enterprises lingered near 10 percent, while tech giants expanded theirs to around 30 percent. Macroeconomic data reinforced this divide, as the Federal Reserve Bank of San Francisco utilization-adjusted total factor productivity metrics showed no clear evidence that artificial intelligence is driving economy-wide productivity improvements.
Usage patterns also underscore this imbalance, with corporate spend management platform Ramp indicating that 99.5 percent of total inference expenditure is concentrated within the top 10 percent of enterprises. Apollo chief economist Torsten Slok pointed out that software firms and startups continue spending on model tokens, whereas traditional companies view inference fees as margin-eroding cost centers rather than engines of productivity. Slok noted that business decisions rely on corporate culture and ethical contexts that calculation-based models cannot fully replace, adding that the success of the technology will depend on whether non-tech firms eventually realize returns. With non-tech chief financial officers facing budgetary decisions over the next six months, enterprise spending cutbacks could soon strain the massive capital expenditure plans supporting the AI infrastructure supply chain.
— AIDEN Editorial Team
What this means for the market
The severe divergence in AI monetization highlights growing risks for the broader technology ecosystem and infrastructure providers. If general enterprises continue to treat model inference as an unprofitable overhead expense rather than a driver of operating margin expansion, demand for advanced hardware and cloud capacity may contract sharply. This dynamic could force hyperscalers and chipmakers to reassess their capital expenditure cycles while directing developer attention toward applications that deliver quantifiable enterprise efficiency.
How this issue is unfolding
Although global technology giants have spearheaded the construction of next-generation AI chips and infrastructure through hundreds of billions of dollars in annual capital expenditures, traditional industries adopting these systems have yet to realize concrete financial returns. According to Apollo analysis, while software developers and startups actively pay token costs for high-performance models, mainstream corporations do not necessarily need marginal performance gains and have begun viewing token fees as a drain on operating expenses. The inherent inability of general-purpose AI to fully replicate contextual business factors, such as corporate culture, accountability, trust, and ethical decision-making, serves as an ongoing obstacle to broader enterprise adoption. Consequently, experts warn that the entire AI infrastructure investment cycle could face sustainability concerns if infrastructure outlays and profit gains remain confined to a handful of technology giants rather than driving economy-wide margins and productivity.
