PwC Research: How Few Firms Capture Most AI Economic Gains

While AI investments break new glass ceilings, research shows that a small group of companies have captured the lion’s share of financial gains while others struggle to move beyond experimentation.

A PwC study interviewing over 1,200 senior executives found that just 20% of organisations are securing nearly three quarters (74%) of AI-driven economic value, underlining a stark divide in how businesses are using the technology.

At the heart of this gap is a shift in mindset. Rather than treating AI as a tool for efficiency alone, leading firms are embedding it into their growth strategies and reshaping how they operate.

“Many companies are busy rolling out AI pilots but only a minority are converting that activity into measurable financial returns,” says Joe Atkinson, Global Chief AI Officer at PwC.

“The leaders stand out because they point AI at growth, not just cost reduction and back that ambition with the foundations that make AI scalable and reliable.”

In addition to Joe, the research was authored by Agnes Koops, Global Vice Chair & Global Chief Commercial Officer at PwC Netherlands and Matt Wood, Global and US Commercial Technology & Innovation Officer (CTIO), Partner at PwC US.

Growth-focused AI strategies

The companies seeing the strongest returns are not simply adopting more tools. Instead, they are rethinking their entire business models with AI at the core. 

“We see it in the data, in the market and in the conversations we have daily with clients from around the globe: AI creates value when it’s aimed at growth, not just cost,” notes Agnes on her LinkedIn. 

“If AI is treated just as an efficiency tool, the biggest opportunity is being left on the table.”

These organisations that lead on AI are 2.6 times more likely to report that they used AI to completely reinvent their business  modeI. 

Businesses that actively use AI to pursue opportunities created by industry convergence are about two to three times more likely to report meaningful financial outcomes from AI than those focused purely on cost savings.

Top performers are also far more willing to redesign workflows from the ground up, than a narrow reliance on productivity. 

Rather than layering AI onto existing systems, they are integrating it deeply into operations, allowing for faster innovation and more flexible decision making.

The research shows that “the most AI-fit companies in our research deliver AI-driven financial performance that’s 7.2 times as high as the other respondents’ performance.”

The study also points out the three measurements of AI use and six foundational capabilities as outlined in the figure that determine AI fitness. 

Automation and trust 

Another defining trait of AI leaders is how they embrace automation, supported by strong governance frameworks. 

These organisations are far more likely to deploy AI in advanced ways, from handling multiple tasks within defined guardrails to operating in increasingly autonomous systems.

This shift is enabling a surge in machine-led decision-making. Leading companies are nearly three times more likely to increase decisions made without human intervention, signalling growing confidence in AI-driven processes.

Crucially, this confidence is underpinned by trust. High-performing organisations are 1.7 times more likely than other companies to invest in responsible AI frameworks and cross-functional governance, ensuring systems operate safely and transparently.

As a result, employees in these companies are significantly more likely to trust AI outputs, accelerating adoption across the business.

A widening divide 

The research points to a growing imbalance between companies that are scaling AI effectively and those still in early-stage experimentation. 

While many businesses continue to test use cases, only a minority are translating these efforts into tangible financial impact.

Without a strategic shift, this gap is expected to widen further. Leading organisations are not only moving faster but are also learning more quickly, refining their models and scaling successful applications at pace.

Success in the AI era therefore depends less on how many tools are deployed and more on how deeply the technology is embedded into growth strategies, decision making and organisational design.