Artificial intelligence has attracted huge investments from businesses, but clear financial gains remain hard to find across much of corporate America. Outside the largest technology companies, there is little evidence that AI has raised revenue, improved profit margins or boosted productivity.
The contrast between major tech firms and other businesses is becoming more important. The seven largest technology companies, known as the Magnificent Seven, have recorded gains in revenue per worker. Yet that measure has remained mostly flat among the other 493 companies in the S&P 500.
Profit margins show a similar divide. The Magnificent Seven have seen their margins rise, while the rest of the index has made little progress. Productivity data also offers few signs that AI is driving stronger growth across the wider economy.
These trends raise a difficult question for business leaders. If AI requires major spending but does not lead to better financial results, how long will companies continue to invest at the same pace?
For many firms, AI costs are becoming harder to overlook. Businesses may pay for model access, computing resources, software and the systems needed to put AI into daily operations. Those costs can weaken profit margins when the tools fail to produce enough value.
Venture-backed companies are spending more on AI tokens as they build products and services around advanced models. Tokens measure the pieces of text and data that an AI system processes. Heavy use can quickly increase bills, especially when companies rely on costly models for large numbers of tasks.
But most traditional businesses may not need the latest model for every job. A cheaper system could be enough to handle routine work, answer basic questions or help employees organize information. Many companies may also be unwilling to pay premium prices without clear proof of a return.
This creates a gap between the excitement surrounding AI and its effect on the wider economy. If the most intensive use of AI agents remains limited to a small group of firms, the technology may have only a modest effect on overall economic growth.
The next test is whether falling AI costs can help mainstream businesses increase sales and improve profits. As models become more widely available, companies may gain access to useful tools at lower prices. However, cheaper intelligence does not automatically create better business outcomes.
A central problem is that business success depends on judgment, not just information processing. AI can calculate, summarize large amounts of data and produce answers quickly. Yet many important decisions require an understanding of people, relationships and the wider effects of a choice.
Companies must consider customer expectations, workplace culture, ethical questions and risks to their reputation. They also rely on unwritten rules that may never appear in a database or formal policy. These factors can shape how customers respond to a product, service or business decision.
Trust and accountability are also essential. A model can suggest an action, but human leaders remain responsible for deciding whether it is appropriate. They must consider who could be harmed, whether a decision is fair and how the company should respond if something goes wrong.
These issues can affect revenue and profit margins over time. A business might save money by automating a task but lose customers if the service becomes less helpful or personal. An AI system may also produce an answer that appears efficient but creates legal, ethical or reputational problems.
This does not mean AI cannot help companies outside the technology sector. It may reduce routine work, support employees and help firms make better use of their data. Some benefits could take time to appear as businesses redesign their operations and learn which tools work best.
Still, companies need more than impressive demonstrations and promises. They need measurable results that show AI is helping them earn more, lower costs or improve the quality of their work. Without those gains, investment may increasingly be treated as an expense rather than a source of growth.
The Magnificent Seven have benefited from their position in the technology industry and their direct role in developing AI systems. The broader S&P 500 faces a different challenge: applying those systems to industries with different costs, customers and business needs.
The future of AI will depend on whether that wider group can turn the technology into lasting financial value. For now, the evidence remains limited. Revenue per worker and profit margins have not shown broad improvement outside major tech firms, and productivity growth has yet to provide a clear sign of an AI-driven shift.
The next phase of the AI boom will therefore be judged not only by more powerful models or larger investment totals, but by whether ordinary businesses can turn the technology into better decisions, stronger revenue and sustainable profits.
