AI value is concentrating. ITPro's report on PwC's 2026 AI Performance Study, a survey of more than 1,200 senior executives across 25 sectors, says three-quarters of AI's financial gains are being captured by 20 percent of companies. Pilots are common. Returns are not.
PwC frames the split around use and operating design, not access to the model itself. The leaders in the study point AI at growth and business-model reinvention rather than only cost reduction. They are also more likely to use advanced applications, autonomous decision-making, responsible AI frameworks, and cross-functional governance boards.
This breaks the pattern of earlier enterprise software. Prior adoption curves handed early movers a temporary lead that eroded as the capability democratized and diffused down-market. The PwC data points the other way, toward a split between companies that orchestrate AI for growth and companies that treat it as a productivity add-on.
The gap is structural, not scale-driven. The leaders in the study are mostly large public companies, which means the result cannot be read as proof that small companies automatically win. The useful lesson for small companies is narrower: the advantage comes from organizational design, not model access.
Small AI-native teams may sit on the favorable side of the split when they design workflows around AI from inception instead of retrofitting operations built for human throughput. That is a hypothesis the PwC data makes plausible, not a conclusion the study proves by company size.
The moat PwC describes forms around organizational design and trust infrastructure, not model access or compute, the two inputs a small company cannot outbid a large one on. The Almanac will watch whether that organizational edge holds once the larger adopters redesign around AI too, or whether the gap PwC measures narrows as the practice diffuses.