59% of finance organizations use AI today, up just one percentage point from the previous year. A figure that signals a significant slowdown after the jump from 37% the year before. GenAI in particular has seen adoption double in the past year, with a 12% increase, but nearly all these new deployments are still confined to pilot programs or early stages, far from full-scale implementations.
The problem is not optimism
63% of CFOs say they are more optimistic about the value of AI in finance compared to a year ago. The optimism is there, but the path has become more complex. Organizations are struggling to identify high-impact AI use cases specific to their own context. This is not a technology problem: it is a problem of applied methodology.
Why spending is not growing at the pace of optimism
Overall investment levels in GenAI remain relatively flat despite growing interest. The most likely explanation is that organizations are deliberately calibrating their spending pace to align with governance requirements and risk management frameworks. In finance, where control, traceability, and auditability are already well-established standards, the most common approach is to test first in areas where these safeguards already exist, before extending AI to less controlled functions.
Where GenAI delivers value in finance today
The areas where GenAI is already demonstrating concrete benefits for finance teams include narrative analysis, unstructured data management, and variance analysis support. GenAI-powered finance chatbots are automating complex activities such as real-time retrieval of financial data, with measurable benefits in reducing manual work. Success in these contexts depends on integration with well-governed datasets and robust security protocols.
AI governance: less popular but strategic
Despite a slight drop in the investment ranking, AI governance maintains overall investment levels that are growing compared to 2025. The emerging pattern is one of closely following innovation with governance guardrails, testing AI frameworks first in finance functions where control and traceability are already established, and then extending them to the rest of the organization with a foundation of real-world experience.
