Credit: Creative Digital Art/Adobe Stock
Finance leaders are willing to devote CapEx dollars to artificial intelligence (AI) only when they can reasonably expect to achieve clear, measurable outcomes. The good news, according to a newly released survey, is many organizations that have invested in agentic AI are seeing a substantial return on investment (ROI).
The research, conducted last October by procure-to-pay solution provider Basware, included feedback from 200 finance professionals across France, Germany, the UK, and the United States. In 2024, a similar Basware study found that finance teams were achieving an average 35 percent ROI on all AI investments. In the 2025 edition, that number is 67 percent.
According to Basware, many of these gains result from finance teams' deployment of agentic AI solutions, which are autonomous or semiautonomous systems designed to perform specific operational tasks on an ongoing basis. The most recent Basware research found that agentic AI is now bringing organizations an 80 percent ROI.
This data comes as nearly half (44%) of finance leaders said they are feeling pressure from corporate management to "do something with AI"—and they are responding. Thirty percent of those surveyed are using AI technologies on a daily basis to automate invoice capture and data entry. Twenty-four percent are leveraging agentic AI daily for cash flow management, 23 percent for scenario modeling and forecasting, and 20 percent to automate financial reporting and reconciliations.
Many of these AI agents are embedded in function-specific software that the company buys, but a full 35 percent of respondents report that they have upgraded financial planning and analysis (FP&A) processes by deploying AI agents they've built in-house using a tool like Microsoft Copilot. When considering the advantages and disadvantages of each approach, Kevin Kamau, director of product management for data and AI at Basware, suggests: "In practice, leaders should buy to accelerate, build to differentiate, and should definitely not do either blindly."
For companies ready to take the leap and hand specific finance workloads off to AI agents, Basware recommends:
1. Develop governance and guardrails. Finance needs to clarify the company's governance policy up front so that AI can take on tasks autonomously without sacrificing accountability. "Choose one process where the impact is easy to measure, define a clear governance model around it, and set decision thresholds so everyone understands when the AI can act and when a human needs to step in," recommends Kamau. "That phased approach … gives finance leaders the confidence that strong guardrails are in place, without slowing their momentum. Governance becomes something that helps you scale safely rather than something that holds you back."
2. Address job displacement. A third (34%) of respondents to the Basware survey indicated that AI agents are already displacing jobs from their department. Finance leaders need to understand, and communicate, ways in which automating time-consuming manual tasks can enable finance staff to take on more value-added, strategic work, rather than simply replacing humans.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.