AI Spending Outpaces Profits as Corporate Returns Remain Unclear

Artificial intelligence investment is expanding across computing infrastructure and corporate operations, but evidence of broad financial returns remains limited. Analyses published by Seeking Alpha on October 23, 2026, RBC Wealth Management on October 22, 2026, and Apollo indicate that investors face contrasting prospects: established technology companies and infrastructure suppliers may benefit from rising demand, while businesses outside the technology sector have yet to demonstrate substantial gains in revenue, margins, or productivity.

A central challenge is the gap between capital spending and cash generation. Seeking Alpha's analysis of AI infrastructure identifies a wide range of financial positions, from hyperscalers financing investment through existing business cash flows to specialized providers relying heavily on debt whose long-term returns remain unproven.

Apollo's analysis finds that revenue per worker and profit margins are rising among the Magnificent Seven technology companies, while both measures remain flat across the S&P 493, the other companies in the S&P 500. The analysis also reports no clear evidence that AI has accelerated productivity growth, raising questions about whether corporate spending will translate into measurable economic benefits.

Apollo argues that mainstream businesses are increasingly treating AI expenses as costs rather than reliable sources of growth. Although venture capital-backed companies are spending more on AI tokens, the analysis says many conventional corporations have little need for the latest models and remain unwilling to pay premium prices for them.

The financial benefits of AI may also depend on capabilities beyond the technology itself. Apollo highlights the importance of judgment, organizational context, accountability, trust, culture, ethical trade-offs, and fairness in determining business outcomes, particularly where decisions affect customer relationships and corporate reputation.

RBC Wealth Management's review of the AI investment boom reaches a related conclusion: access to AI alone does not guarantee a lasting competitive advantage. Companies with proprietary datasets, patents, established customer relationships, and other difficult-to-replicate assets may be better positioned to turn the technology into durable commercial value.

The firm points to subscription software businesses that process customers' own data as an area facing greater competitive pressure. As AI makes software development easier, barriers to entry can decline, potentially weakening the profitability of businesses whose services are relatively easy to replicate.

RBC also highlights RELX as a contrasting example. Rather than relying solely on AI as a standalone product, the company has partnered with Anthropic, using its proprietary, legally protected, and difficult-to-replicate data resources alongside AI technology. The example illustrates how established information assets can support the adoption of new tools.

Competition among AI developers presents another challenge for investors. RBC describes an industry divided between frontier laboratories seeking technological leadership and open-weight model developers offering alternatives at a fraction of the cost. If companies continue spending aggressively to avoid falling behind, competition could limit their ability to sustain high profit margins.

RBC compares this dynamic with the solar energy industry, where transformative technology and intense competition contributed to prolonged periods of weak investment returns. The comparison underscores the distinction between technological progress and the ability of businesses to capture lasting profits from it.

Infrastructure suppliers have so far occupied a more favorable position in the AI expansion. RBC identifies semiconductors, electricity, cooling systems, backup power, and electrical components as essential inputs for data center construction. These suppliers can earn revenue while technology companies continue expanding computing capacity, regardless of which AI applications ultimately dominate.

However, infrastructure exposure does not eliminate financial risk. RBC notes that semiconductor markets remain cyclical and that suppliers' fortunes depend partly on continued spending by AI developers. The analysis also cites a reported loss of close to $5 billion for SpaceX in 2025 and says meaningful profitability for AI laboratories remains uncertain.

The investment outlook therefore differs across the AI ecosystem. Cash-rich technology companies can finance infrastructure from established operations, while highly leveraged providers face greater uncertainty about whether future earnings will justify their spending. For non-tech corporations, the unresolved question is whether increasingly affordable AI tools can generate enough additional revenue or cost savings to improve financial performance.