Three quarters of enterprise finance leaders plan to increase AI investment over the next 12 to 24 months, but only 39% have built the operating model to scale it, according to a new Forrester Consulting study commissioned by Basware. The findings, drawn from 231 finance and AP decision-makers across the US, UK, France, and Germany, reveal that the era of experimental AI spending is giving way to a harder standard: prove the return before you get the budget.
68% of respondents said they require demonstrable ROI before committing further technology funds. The shift reflects where AI now operates in finance - not in pilots, but in live accounts payable workflows where process complexity quickly exposes weak business cases.
Investment rises alongside expectations
Finance leaders have moved past the promise of instant AI returns. Only 7% expect AI investments in AP to pay back in under six months. A larger group - 20% - anticipates payback within six to 12 months, while 35% expect value to take 13 to 24 months to materialize. The demand for proof is tightening as regulatory pressure mounts. 65% of respondents need major or urgent improvement to meet new financial regulations, including Nacha's 2026 fraud monitoring rules now in effect in the US.
67% of finance teams already run AI for targeted AP use cases. But running AI in pockets is not the same as scaling it. Only 39% operate AI centers of excellence at scale. "Finance is a strong place to start with AI because the value can be measured," said Donna Wilczek, Chief Product and Technology Officer at Basware. "The challenge is getting from ambition to execution in a way the business can trust. Once outcomes are proven, the remit can grow."
Governance becomes the deciding factor
Control now outweighs raw innovation in purchasing decisions. 64% of finance leaders prioritize stability and compliance when choosing AI tools. Less than half - 46% - say they have struck an effective balance between governance and innovation. The tension is clear: AI can resolve invoice exceptions and recommend approvals, but it needs permission to act and a reliable record of what happened.
For CFOs and finance chiefs building their AI Learning Path for CFOs, the study signals that governance frameworks are no longer optional. "Governed AI is no longer aspirational, it's a board-level requirement," Wilczek said. "Every AI decision in accounts payable needs to be logged, traceable, and auditable from the moment it's made, not reconstructed after the fact."
From automation to trusted execution
Basware's response to the governance gap is a framework called Governed Autonomy, which sets three levels of AI authority: Advisor, Collaborator, and Operator. Human review stays in the workflow wherever judgment is required. AI's authority expands only as outcomes are proven at each level. The approach reflects a broader shift in AI for Finance, where AP has become an early test of whether AI can move from automation to trusted, auditable execution.
AI investment is concentrating in AP areas where the link to control is easiest to demonstrate. 55% of respondents expect AI investments in AP to pay back within six to 24 months, a timeframe that aligns with the governed, incremental approach the framework describes. "Success in the next phase of AP won't be achieved by the teams using the most AI, but by the teams that govern it best," Wilczek added.
Why this matters for finance leaders
The study makes one point clear: the barrier to AI adoption in finance is no longer the technology. It is the operating model. With 68% of leaders requiring proven ROI before further spend, finance teams need to build the governance structures, audit trails, and compliance controls that let AI scale safely. The 39% who have already built centers of excellence at scale are pulling ahead - not because they invested more, but because they built the scaffolding to prove value in live operations.
Your membership also unlocks:
