Most Companies Still Aren't Winning With AI, and the Gap Is Growing

There's a stat from a new PwC study that every business leader needs to see.

74% of the economic value being generated by AI is going to just 20% of companies. The other 80% of organizations are splitting the remaining 26%.

That's not a projection or a think piece. That's the finding from PwC's 2026 AI Performance Study, released April 13, based on interviews with 1,217 senior executives at large companies across 25 sectors worldwide.

And if that doesn't get your attention, this might: the top-performing group is generating 7.2 times more AI-driven value than the average competitor. They're also showing profit margins 4 percentage points higher.

Why the gap exists

Here's what's interesting. The difference isn't about who's spending the most on AI tools. It's about what they're pointing AI at.

Most companies are using AI to cut costs. The leaders are using it to grow revenue. That's the core split.

Specifically, the top 20% are 2.6 times more likely to say AI is helping them reinvent their business model. They're two to three times more likely to use AI to find growth opportunities across industry boundaries, what PwC calls "industry convergence." That factor turned out to be the single strongest predictor of AI-driven financial performance in the study.

Think of it this way: most companies bought AI tools and layered them on top of existing processes. The leaders redesigned how their businesses work around what AI makes possible.

The pilot trap

Here's the pattern PwC keeps finding, and it's consistent with other data too.

A lot of companies are running AI pilots. They look busy. They have dashboards and demos. But most of those pilots never scale into the actual business.

The study found that 56% of companies have seen no significant financial benefit from AI to date. Only 12% reported gains in both cost and revenue. And CEO confidence in near-term revenue growth dropped to a five-year low. Just 30% feel good about the next 12 months, down from 38% in 2025.

This lines up with MIT research from 2025 that found 95% of enterprises reported zero return on generative AI pilots. The technology works. The implementation doesn't.

PwC Global Chairman Mohamed Kande put it pretty directly at Davos earlier this year: 2026 is the year where a small group of companies are turning AI into measurable financial returns, while most others are still stuck trying to get past the pilot stage.

What the leaders do differently

PwC's study looked at 60 different AI management practices and grouped them into what they call an "AI fitness index." A few things stood out about the top performers.

They're nearly twice as likely to run AI in advanced ways. Not just answering questions, but handling multiple tasks within defined guardrails or operating in self-optimizing modes. They're increasing the number of decisions made without human intervention at nearly three times the rate of their peers. And they've built the governance structures to do this safely, with responsible AI frameworks and cross-functional oversight boards.

The other finding worth noting: technology itself accounts for only about 20% of the value in a typical AI initiative, according to PwC. The remaining 80% comes from redesigning workflows so that AI handles routine work and people focus on higher-value tasks.

Most companies invest heavily in the 20%. The models, the APIs, the tools. They skip the 80%.

Why this matters now

This isn't one of those "the future is coming" stories. The divide is already here, and PwC is saying it's going to get worse.

Companies that have built the foundations (good data, clear governance, workflow redesign, a focus on growth over cost-cutting) are compounding their advantage. They learn faster, scale what works, and automate decisions safely.

Companies that haven't? They're not just behind. They're falling further behind with every quarter that passes.

PwC's warning was direct: the gap will widen quickly for those that don't act.

If your AI strategy is still mostly about "saving time" or "being more efficient," this study is a signal that the bar has moved. The companies capturing real value from AI aren't just doing things faster. They're doing different things entirely.

That's the real story here.