Four out of five Chief Financial Officers (78%) have expressed a desire to increase their investment in AI over the next 12 to 18 months, recognising its value in transforming financial operations, but admit they ‘don’t know where to start’, according to a new global survey.

While this signals growing confidence in AI, it’s accompanied by a strict demand for accountability. Half of CFOs (50%) report they will axe AI investment if it fails to deliver measurable ROI within a year, highlighting a narrow window for proving its impact.

The findings were revealed in accounts payable automation Basware’s AI to ROI Report, conducted by independent research firm Financial Times Longitude. The study surveyed 400 global CFOs and finance leaders on their attitude, priorities and challenges when it comes to AI in the finance department.

The biggest priority for financial leaders when implementing finance transformation projects is cost efficiency, with one-third (32%) listing it as their primary focus.

However, any financial confidence will remain rocky with macroeconomic and geopolitical uncertainties continuing to swirl up until the end of the year, and beyond.

This uncertainty is directly impacting AI adoption within finance functions, making it challenging for 41% of finance leaders to prioritise AI investment.

No one is exempt from the fallout of uncertainty. Even Meta’s latest earnings saw a reduced stock-price revision worth $5.1bn to reflect concerns over the company’s capacity to mobilise AI investments and sustain growth.

Meta has invested in a “significant acceleration” in AI-related infrastructure, but analysts warn that they need to prove they can continue to cover its rising AI costs and deliver ROI. For companies like Meta, a bounce-back is likely, but recoverability in the short-term isn’t achievable for all.

According to the report, finance leaders want to invest more in AI to reduce manual tasks and allow their teams to focus on priorities in their business. 70% of finance leaders said that staff want AI support for administrative tasks, while 75% report that AI has enabled their workforce to focus on more strategic activities such as e-invoicing compliance.

However, the biggest barrier to finance transformation is change management and an unclear AI strategy. 40% of finance leaders state that their organisation lacks change management capabilities, while 31% highlight a lack of clear strategic vision for the future of the finance function.

These factors hinder ROI for AI projects, resulting in question marks over investment without tangible financial benefits.

Perttu Nihti, Chief Product Officer at Basware, commented on the new AI study: “The office of the CFO is tasked with overseeing a complex range of functions from regulatory compliance through to cash flow management and financial reporting. All of which are areas where AI-powered automation can help to reduce hours and relieve pressure. But the success of AI investment hangs on knowing where to start and proving impact. We’re at the AI tipping point. Focusing on high-value wins, such as AI-powered efficiency that demonstrate quantifiable ROI quickly, in areas such as compliance, error reduction and fraud detection, will help justify investment across a company’s organisation.”

Accounts Payable is one such starting point that is already returning greater levels of ROI. The survey revealed that organisations prioritising AI investment are realising a return of 136% ROI, with savings exceeding $1.36 for every $1m invested over the duration of three years.

Among CFOs prioritising AI innovation and applying it to processes in accounts payable, reducing errors, faster fraud detection, reduced operational costs and improved regulatory compliance were found to be the greatest benefits.

One example of how AI has driven ROI for an enterprise is paper and packaging manufacturer Billerud. The accounts payable team was previously hindered by time-consuming manual invoice processing, which cost them several hours each day. To address this inefficiency, they implemented Basware’s SmartPDF AI Instant Learning solution, which uses AI and text extraction to convert PDF invoices into authentic e-invoices.

Billerud has seen a substantial reduction of invoices needing validation, dropping from 15% to 9% due to AI Instant Learning, and saving AP staff several hours each day. Now, powered by AI, over 90% of their invoices are validated automatically due to faster, more accurate data extraction.

Jesper Persson, Business Developer at Billerud, said: “Since day one, we’ve perceived the desired values from the project. The quality of invoices has improved considerably, and the AI continues to evolve and improve with each passing day. The efficiency gains we achieved translated directly into tangible cost savings, paving the way for a rapid return on investment within just a few months.”

The findings reveal a tipping point for AI in finance. While 50% of CFOs may cut AI investments without clear ROI, those that identify specific starting points for AI implementation – such as invoice processing automation – can achieve measurable returns within months rather than years. These successes are creating a blueprint for broader AI adoption across finance functions.