Finance chiefs demand proof before boosting AI spend

A Forrester Consulting study commissioned by Basware found that only 7% of finance leaders expect artificial intelligence in accounts payable to deliver payback within six months. At the same time, 76% plan to increase AI investment.

The findings suggest a more cautious approach to AI spending in finance teams. Some 68% of respondents said they want a demonstrable return on investment before committing more technology budget. The research covered 231 decision-makers responsible for finance and accounts payable technology across the UK, US, France and Germany.

While investment plans remain strong, many organisations appear to be at an early stage in putting in place the controls needed for wider AI use. The survey found that 67% of finance teams already use AI in targeted accounts payable tasks, but only 39% operate AI centres of excellence at scale.

That gap between adoption and operational readiness is reflected elsewhere in the results. Nearly half of those surveyed, 46%, said they had struck an effective balance between governance and innovation, while 64% said they prioritise stability and compliance over innovation when selecting AI tools.

Measured returns

Expectations for financial returns were spread over a longer timetable than much of the recent debate around AI has implied. In addition to the 7% who expect payback in less than six months, 20% said they expect returns within six to 12 months, and 35% said value would take 13 to 24 months to emerge.

The report places accounts payable at the centre of that discussion because finance teams can track outcomes in a relatively structured environment. It suggests leaders are using those processes to test whether AI can work within day-to-day systems rather than as a stand-alone experimental tool.

Regulatory pressure is also shaping those decisions. Some 65% of respondents said major or urgent improvement is needed to adapt to new financial regulations, while 63% cited growing demand for data-backed decision-making.

Against that backdrop, organisations appear to be weighing automation against oversight. Basware argued that any AI system handling invoice exceptions or approval recommendations needs clear permissions and a record of its actions.

Donna Wilczek, Chief Product and Technology Officer at Basware, said finance departments are a practical setting in which to test the business case for AI.

"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."

Control question

The survey indicates that trust and control are now central issues for companies moving from limited automation to broader AI use in operational workflows. Finance teams want systems that can be governed closely, especially where decisions have audit, compliance and payment implications.

That concern is reflected in Basware's framing of "governed autonomy", under which AI would take on more authority only after results have been tested. Human review, the company said, should remain in workflows where judgement is still required.

Wilczek said the need for oversight has moved beyond a technical preference and become a broader management issue.

"Governed AI is no longer aspirational, it's a board-level requirement," she 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."

The research comes as companies face sharper scrutiny over how autonomous systems behave in live environments, particularly when those systems are given scope to act within business processes. In that context, finance functions offer a clear example of both AI's appeal and the constraints that may limit wider deployment.

Basware said its framework for AI in accounts payable sets out three levels of authority, moving from advisory roles to greater operational responsibility as confidence grows. The broader message from the survey, however, is that many finance leaders are not yet ready to move quickly without firmer evidence that the systems can deliver returns and remain under control.

"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," said Wilczek.