Most companies using artificial intelligence are not making money from it. A small group is pulling away from everyone else, and the gap is growing.
That is the core finding of PwC’s 2026 AI Performance Study, published April 13. The global study surveyed 1,217 senior executives at large, publicly listed companies across 25 sectors, measuring the revenue and efficiency gains attributable to AI.
It found that 74 percent of AI’s economic value is captured by just 20 percent of organizations.
The remaining 80 percent of businesses share the other 26 percent.
What Separates the Leaders
The top-performing 20 percent of companies generate 7.2 times more AI-driven revenue and efficiency gains than the average competitor. The divide is not primarily about how much AI these companies deploy but about what they direct AI toward.
The leading companies are approximately two to three times more likely to use AI to identify and pursue growth opportunities and reinvent their business model. They are twice as likely to redesign workflows to incorporate AI rather than simply adding AI tools on top of existing processes.
Automation at Scale
AI leaders are increasing the number of decisions made without human intervention at almost three times the rate of their peers. Companies with the best AI-driven outcomes are 1.8 times more likely to use AI to execute multiple tasks within guardrails and 1.9 times more likely to operate AI in autonomous, self-optimizing ways.
In practice this does not mean AI operates without oversight. It means routine decisions that previously required human approval – pricing adjustments, inventory management, fraud detection, customer segmentation – now execute automatically within defined boundaries. Humans set the rules and monitor outcomes. They no longer approve every individual action. The intervention moves from the transaction level to the strategy level.
That automation gap compounds over time. The more decisions a company automates reliably, the faster it learns, the faster it scales, and the harder it becomes for slower organizations to close the distance.
CEO Confidence Is at a Five-Year Low
The PwC data connects directly to a broader economic concern. Only 12 percent of surveyed companies have seen AI deliver both cost and revenue gains. CEO confidence in revenue growth fell to a five-year low, with just 30 percent expressing confidence about their company’s revenue prospects for the next 12 months, down from 38 percent in 2025 and 56 percent in 2022.
Most business leaders are spending on AI without seeing returns. Not because the technology does not work, but because most AI activity stays in pilots that never reach the core business.
The Convergence Advantage
PwC’s analysis identified capturing growth opportunities from industry convergence as the single strongest factor influencing AI-driven financial performance, ahead of efficiency gains alone.
The leaders are not just using AI inside their existing business model. They are using it to move into adjacent markets and create revenue streams that did not exist before. That is a fundamentally different strategy from cost reduction, and it is producing fundamentally different results.
The Warning
PwC warned directly: without a shift in approach, the performance gap between AI leaders and laggards is likely to widen further as leading companies continue to learn faster, scale proven use cases, and automate decisions safely at scale.
The 74/20 split is not a temporary imbalance waiting to correct itself. It is a structural divide that favors organizations that moved early, built the right foundations, and pointed AI at growth rather than efficiency. For the majority still in pilot mode, the window to close that gap is narrowing with every quarter.