Only 7% of Dow 30 companies quantified an AI-linked business result in their 2025 annual filings, even though 77% described actively using the technology.
Flexera, a technology cost-management software company, reviewed the annual filings of all 30 companies in the Dow Jones Industrial Average. Its analysis of corporate AI reporting found that 90% identified company-specific AI risks, while just 7% disclosed an identifiable AI investment amount.
Overall, 93% of the companies discussed AI in their filings and 73% identified a specific application.
The analysis covers public disclosure, not actual returns. It does not establish whether companies are failing to earn money from AI or whether they measure the results internally.
A similar gap in UK reporting
Flexera found a similar reporting gap in the U.K. Its separate review of the FTSE 100, an index of major London-listed companies, found 36% described substantive AI use and 32% linked AI to a specific business outcome. Yet only 3% reported a specific financial outcome and 2% disclosed an AI spending figure.
Marlon Oliver, Flexera’s senior vice president for Europe, the Middle East and Africa, said, “AI investment needs to connect to business KPIs, whether that is downtime reduced, tickets resolved, claims processed or revenue generated.”
Oliver also argued that shareholders should be able to see what those improvements are worth, what they cost and whether the economics are improving. An operational result alone does not establish a financial return.
Putting a price on AI use can also be difficult. Flexera said AI capabilities may arrive through existing software subscriptions or applications built on cloud infrastructure. The team benefiting from an application may not be the one paying for it. For AI agents, each additional action can consume more tokens, data and computing power, said Steve Daheb, Flexera’s head of AI go-to-market strategy.
McKinsey finds limited earnings impact
McKinsey, a management consulting firm, examined AI returns in its August 2026 global survey, “The state of AI in 2026: On the road to ROI.” Among 1,719 respondents, 37% attributed a positive impact on earnings before interest and taxes (EBIT) to AI. About 6% attributed at least 5% of EBIT to AI and reported significant value.
The survey also found that one in five respondents said operating costs, including tokens, had constrained AI use. Unlike Flexera’s analysis of public filings, McKinsey’s findings were based on respondents’ assessments of their organizations.