PwC Surveyed 4,454 CEOs Across 95 Countries. 56% Say AI Hasn't Paid Off Yet. Only 12% Got Both Revenue and Cost Gains.

November 11, 2026 · Josh Agusti, Founder, Aiven Consulting Group

PwC's 29th Global CEO Survey polled 4,454 chief executives across 95 countries and territories, and the picture it paints of corporate AI spending is more sober than the rhetoric surrounding it.

Fifty six percent of CEOs say their company's use of AI has not yet produced either higher revenue or lower costs. Only twelve percent say it has delivered both. Enterprise AI budgets keep climbing. For most of the executives funding them, the payoff has not yet shown up on a financial statement.

The Math Behind the Headline

Break the finding down and a more specific shape appears. Roughly three in ten CEOs report higher revenue from AI over the past year. About one in four report lower costs. A further third see some benefit on one side or the other. What is left, at more than half the sample, is a group that has funded AI capability through two or three budget cycles and still cannot point to a number that moved. For a board asking what that spending actually bought, pilot counts and roadmap slides are no longer a satisfying answer, which is part of why more finance and strategy teams are turning to a disciplined read on AI ROI measurement instead of another status update.

The pattern isn't confined to one firm's numbers either. A separate global study from KPMG on AI investment scaling tracked AI spending intentions through the current economic cycle. Most of the executives surveyed there plan to keep raising their AI budgets regardless of the downturn. Far fewer of them, by their own account, have gotten AI running at scale across the business.

Why Deployment Isn't the Same as Value

The more useful reading of PwC's numbers isn't that the technology underperforms. It's that most organizations have deployed AI inside workflows that were never redesigned to use it well. A drafting tool layered on top of an approval chain that still routes through six people, a chatbot bolted onto a support queue that hasn't otherwise changed: these generate activity without generating a number finance can defend. The CEOs who report real movement in revenue or cost describe a different starting point. They treated AI as a reason to rebuild how a team actually works, not as a feature added to how the team already worked. That's a strategy and operating model question before it's a technology one, and it's closer to what redesigning the operating model around AI actually requires than another round of tool evaluation.

This is also why the gap tends to persist even as budgets rise. Adding a better model or a faster agent to an unchanged process rarely clears the bar investors expect; it just makes the unchanged process marginally faster. The organizations showing up in the twelve percent, the ones getting both revenue and cost benefit, typically changed who makes a decision, how a handoff happens, or what a team is accountable for, and then layered AI into that new shape of work.

What CEOs Say When the Number Doesn't Move

PwC's respondents weren't uniformly bearish. A meaningful share expect their AI investment to climb again in the next cycle regardless of this year's returns, betting that today's missing ROI reflects the early, awkward phase of a long transition rather than a ceiling. That bet isn't unreasonable. Enterprise technology shifts of this scale rarely show clean financial returns in their first two or three years, and a CEO abandoning AI spending now would be trading a slow, uncertain payoff for a faster, certain cost of falling behind competitors who stay the course. The more defensible position sits somewhere in between: keep investing, but stop treating the investment as self-justifying.

Boards reading this survey are unlikely to stay patient with another year of pilot updates. Fifty six percent without upside is a number every CEO in that seat has either already explained internally or will be asked to explain soon. The ones with a credible answer will be the ones who can show that the first year or two of AI spending went toward rebuilding how work gets done, not just toward buying more of the technology itself.

Key Takeaways

PwC's 29th Global CEO Survey of 4,454 CEOs across 95 countries found that 56 percent have not yet seen either higher revenue or lower costs from their company's AI investment, and only 12 percent report gains on both fronts.
The CEOs reporting real financial returns are disproportionately the ones who redesigned how a team or process works around AI, rather than layering AI tools onto an unchanged workflow.
Closing the ROI gap starts with measuring AI's actual financial impact against a full budget cycle, not a favorable pilot quarter, and building that discipline into how AI investment decisions get made.