The story of artificial intelligence in finance has shifted from whether companies should adopt it to whether they can actually make it work at scale. A new global study of 1,013 senior finance leaders across 20 countries found that while more than three-quarters of organizations are now using AI in financial planning, reporting, and commercial analysis, only 23 percent report that AI is exceeding their expectations. This gap between broad adoption and exceptional performance reveals a critical truth: having AI and using it effectively are two entirely different challenges.

Why Are Most Finance Teams Adopting AI But Not Seeing Results?

The research, conducted by KPMG International in March 2026, shows that 71 percent of finance organizations report their AI investments are meeting or exceeding return on investment (ROI) expectations. Yet when asked specifically whether AI is exceeding expectations, only 23 percent agree. This apparent contradiction points to a deeper problem: adoption is moving faster than the operating capability to translate it into enterprise-wide performance at scale.

The organizations capturing real value from AI are not adopting more of it. Instead, they are directing it at specific types of work where it matters most. Finance leaders who deployed what researchers call "agentic AI" (AI systems that can take autonomous actions within defined parameters) reported gains that separated them from the rest by an average of 32 percentage points, growing to nearly 40 percentage points on forecast accuracy and ROI. This suggests that strategic deployment beats volume deployment every time.

Where Is AI Actually Delivering the Biggest Wins in Finance?

The strongest gains from AI in finance are not coming from automating routine, transactional tasks. Instead, AI is excelling at judgment-heavy work, precisely where finance teams have historically struggled most. Decision-making quality improved for 70 percent of organizations using AI, decision-making speed improved for 71 percent, and forecasting accuracy improved for 64 percent. These are the decisions that require interpretation, context, and nuance, not just speed.

This finding challenges a common assumption about AI in business: that its primary value lies in replacing human effort on repetitive tasks. In finance, the real leverage comes from augmenting human judgment, not replacing it. Organizations are using AI to sharpen their ability to forecast, interpret data, and make faster, more confident decisions in uncertain conditions.

How to Build an AI Operating Model That Actually Delivers Results

The research identifies four reinforcing priorities that separate leaders from the rest:

• Reframe AI Around Value, Not Tasks: Stop counting AI implementations and start measuring what decisions improve and what business outcomes change. Leaders are directing AI at the work where judgment matters most, not just automating what is easiest.
• Treat AI Governance as the Ticket to Play: Organizations that can produce AI audit evidence efficiently report three to six times the rate of significant improvement compared to those that cannot. On error reduction, 33 percent of organizations with strong assurance readiness reported significant improvement versus only 6 percent without it. On confidence in scaling, the gap was 42 percent versus 14 percent.
• Build Measurement Into Execution: Assurance readiness is a stronger predictor of performance than key performance indicator (KPI) tracking alone. Governance is often framed as a brake on AI adoption, but the data shows the opposite: it is the accelerator.
• Shape the Total Workforce, Not Just Training: Most organizations are upskilling existing finance teams (38 percent), but only 28 percent are hiring for different skillsets. The constraint is not just capability; it is the composition of the team itself.

These four priorities form a reinforcing cycle. Decision-oriented AI compounds with governance; governance scales with measurement; measurement translates into action only with the right workforce. Built together, they create what KPMG researchers call the "Decision Advantage".

"The conversation about AI in finance has changed. Two years ago, the question was whether AI could deliver. Today the question is what it should be deployed to do. The organizations getting it right are using AI to sharpen judgment, not just to speed up tasks," said Sebastian Stöckle, Global Head of Audit Innovation and AI at KPMG International.

Sebastian Stöckle, Global Head of Audit Innovation and AI, KPMG International

What Is Holding Finance Teams Back From Scaling AI?

Two critical constraints emerged from the research. First, data quality remains both the most cited barrier and the most cited opportunity. Thirty-six percent of organizations identified improving data quality, integration, and system interoperability as their greatest opportunity to extract more value from AI in finance, yet many also named these as significant vulnerabilities. The constraint is not the technology itself; it is the condition of the data that AI depends on.

Second, workforce capability is a distinct challenge requiring its own response. Data fluency, defined as the ability to assess data quality, interpret AI outputs, and communicate findings the business can act on, is the most critical capability need. It sits at the intersection of finance expertise and AI literacy. The leaders are doing both: upskilling existing teams while hiring for a different orientation to data.

"Adoption is no longer the differentiator. The leaders are building the operating conditions around AI, governance, measurement, and a workforce equipped to act on what AI produces. Trust, embedded in how performance gets built, is what separates the organizations capturing value from the rest," noted Nikki McAllen, AI in Financial Reporting and Audit at KPMG.

Nikki McAllen, AI in Financial Reporting and Audit, KPMG

The 2026 Global AI in Finance report reveals that the next phase of enterprise AI is not about adoption breadth; it is about operating discipline. Finance leaders who want to move from pilot projects to sustained, scaled value must shift their focus from how much AI they deploy to how well they govern it, measure it, and equip their teams to act on it. Trust, operationalized through governance and controls, is becoming the defining competitive advantage in finance.