Insights

Finance functions have more than doubled their use of AI in two years, yet fewer than one in four companies has achieved results that exceed expectations. According to KPMG International's third survey of AI adoption, which covered 1,013 finance executives across 20 countries, the proportion of companies actively using AI in financial planning, reporting, and commercial analysis surged from 30% in 2024 to 75% in 2026, taking adoption into the mainstream. Yet only 23% said the return on AI investment had exceeded expectations, revealing a clear gap between the pace of adoption and the ability to translate it into tangible performance.

From a cost-reduction tool to a decision engine. AI delivers its strongest results in judgment-intensive work rather than simple transaction processing. Respondents cited improvements in decision-making speed (71%), decision quality (70%), and forecasting accuracy (64%). Companies that have adopted agentic AI outperformed non-adopters by an average of 32 percentage points across key measures including close efficiency, forecasting accuracy, and ROI, with gaps of approximately 40 percentage points in forecasting accuracy and ROI. Use is also expanding into strategic value creation, including greater capacity for growth (36%) and improved customer experience (35%). Agentic AI is already moving beyond initial adoption toward orchestration, in which multiple AI tools are coordinated autonomously.

Data and governance determine the performance gap. The gap in AI outcomes across industries reaches as much as 29 percentage points. Banking, supported by regulation-driven data accuracy, leads on five of six measures, while healthcare and life sciences rank last. Companies that track AI-related KPIs report a 10-percentage-point higher rate of ROI improvement than those that do not (68% versus 58%), and companies capable of producing AI audit evidence show improvement rates three to six times higher than those that cannot. Companies with AI governance also report a 36% reduction in operational errors, far ahead of the 6% recorded by unprepared companies.

Implications for M&A. First, because data quality, integration, and interoperability are identified as the greatest opportunities for creating value from AI, demand is likely to grow for acquisitions of and strategic collaboration with finance data infrastructure and AI governance solution providers. Second, agentic AI capabilities in finance can increase due diligence speed and forecasting accuracy, improving the efficiency of the M&A deal process itself. Third, valuation differentiation is likely to become more pronounced between companies with proven AI results and those without them, making AI operating maturity a due diligence factor that influences deal premiums.