Survey Finds Financial Firms Lead AI Adoption Across 27 of 75 Tasks

Revenue recognition and accounting close was the most widely adopted financial-services use case at 65%. Credit risk assessment and sales forecasting each reached 60%. PYMNTS characterized these as structured workflows where firms can test results and trace how decisions were made.

Customer-facing uses were less mature in the survey:

• •Churn prediction: 30%
• •Know Your Customer identity verification: 20%
• •A/B testing and experimentation: 10%

The contrast suggests that the surveyed firms were concentrating AI in established internal processes before applying it broadly to personalization and customer acquisition. That is an interpretation of the reported distribution, not a direct measure of model quality or business value.

PYMNTS reported that 85% of financial-services respondents expected to increase AI budgets over the following 12 months. Productivity and competitive positioning were each cited by 65% as investment reasons, while 55% cited risk reduction and compliance.

For data and ML teams, the task-level results point toward familiar production requirements: reliable lineage, validation, monitoring, access controls, and auditability, especially where outputs affect financial reporting or lending decisions. The survey does not establish how effective those deployments are or whether they are fully automated.

The figures should be read as a directional view of reported adoption among large US enterprises. With only 20 respondents per sector and one executive representing each company, the results are not a census of the financial industry and should not be generalized to smaller firms or other markets without additional evidence.