Report

From Experimentation to Execution: Governed Autonomy for AI in Accounts Payable

Download the study

67% of finance teams already run AI in accounts payable, in live, day-to-day work. The question has shifted, from whether to adopt AI, to how best to control it. This Forrester Consulting Opportunity Snapshot, commissioned by Basware, asked 231 enterprise finance leaders across the UK, US, France, and Germany how they are moving AI in AP from experimentation to execution, and where the gaps still sit. One pattern stands out: the teams pulling ahead are not the ones using the most AI. They are the ones governing it best. They set how far the AI can act, keep a human in the decision loop, and hold an audit trail that defends every call. Basware refers to this approach as Governed Autonomy.

Inside, the data shows where AI in AP already pays off, where execution stalls, and what it takes to prove the return.

The state of play
Where AI pays off
The execution gap
Govern it well
Download the study

AI in AP is already running

Have moved past pilots to AI in live, targeted AP use cases.

Control decides what ships

Prioritize stability and compliance over raw innovation when they choose AI.

Regulation raises the bar

Need major or urgent improvement to meet new financial regulations.

Top 3 AP functionalities that would most benefit from AI

Where does AI in AP already pay off?

Finance teams are not spreading AI thin. They aim it at the AP functions where data is strong and the benefits of automated invoice processing are clearest.

Invoice data capture leads, the top function to benefit from AI (43%). This is where AI invoice automation pays off first.
Invoice matching is close behind (30%), the other high-volume, rules-based task automated invoice processing software handles well.
Fraud detection and risk management follow (28%), where accounts payable fraud detection and fraud prevention add control, not just speed.

Why is execution the gap, not ambition?

The foundations are in place. Over 60% confirm clear AI guidelines, leadership support, and risk frameworks. The same holds across AI in accounting and finance: strategy is not the gap. The operating model is. Only 46% balance governance with innovation, and just 39% run an AI Center of Excellence. The structure exists on paper but is not yet built into the AP automation workflow. That is the case for AI powered AP automation that is governed from the start, not bolted on.

How organizations score across the four pillars of AI readiness

Principles and strategy

Governance and process

Operating model

Technology and data

How much are organizations planning to increase AI investment?

Organizations Increase AI Investment With Expectations Of Measurable Returns Within 24 Months

Note: Figures shown represent the majority of responses; a small number of outliers have been excluded.

What does it take to win the next round of AI funding?

Budgets are opening. 76% plan to increase AI investment, but 68% want demonstrable ROI first. The most valuable AI applications in finance are the ones teams can prove.

Expectations are realistic: over half expect payback in 6 to 24 months, 35% at 13 to 24 months, only 7% under 6 months. This is a medium-term play. That makes defensibility the product: show returns early, keep showing them, and put an audit trail behind every decision.

Three ways to turn AI ambition into governed execution

Set how far the AI goes

Configure autonomy by use case: Advisor, Collaborator, or Operator, with a human in the loop at every stage. Control is a setting, not a leap of faith.

Build the audit trail in from the start

Log every AI decision so it is traceable, explainable, and defensible to an auditor or regulator. Compliant by design beats reconstructed after the fact.

Make governance part of how you work

Move from guidelines and good intentions to an operating model that runs day to day. That is what closes the gap between AI strategy and AI at scale.

Autonomy and integrity were never a trade-off. The whole job is refusing to choose between them.

Donna Wilczek, Chief Product & Technology Officer, Basware

Submit the form below to download the resource

Get the full Forrester Opportunity Snapshot, including all data, charts, and the methodology behind these findings on AI in accounts payable. If you own finance, AP, or AI strategy at a large enterprise, this report is for you.

Base: 231 enterprise decision-makers and finance professionals responsible for finance/AP
Source: Forrester’s Q1 2026 AI In Accounts Payable Automation Survey [E-65952]