CFOs are embracing AI but want stronger data, controls and accountability first, with practical use cases in forecasting, reconciliation and payments leading adoption.
Finance leaders are under pressure to move faster on AI, but recent research suggests CFOs are not entirely prepared to trade speed for control, confidence, or accountability.
Bottomline research conducted by Censuswide among 414 CFOs in UK and US companies with more than 500 employees points to a measured reality: AI is rising quickly on the finance agenda, but many teams still need better data, systems and controls before they feel they can use it safely.
The strongest signal is not resistance. Almost half of CFOs describe their finance function as ready to adopt AI, while 43% see its value but are limited by current foundations and concerns about safe scaling.
Only one in ten say they are cautious because of accuracy, risk and compliance. CFOs are not rejecting AI; they are asking whether the data, systems and controls around it are mature enough to support financial, operational and regulatory decisions.
Cash visibility and forecasting remain practical barriers
Cash visibility problems are well known to the CFO office. Forecasting still relies on spreadsheets or manual downloads. Exceptions and reconciliations require too much manual effort. Data is spread across multiple ERP or finance systems, and finance and operational data cannot easily be combined. These barriers directly affect the ability to forecast, report, and act with confidence.
Fewer than half of CFOs responding said they are very confident in their ability to forecast cash accurately over the next 30 days. Confidence falls to 44% at 60 days and 41% at 90 days, creating a clear gap between expectation and capability.
Manual work is a key area where core processes could benefit from AI. That includes board reporting, duplicate payment or error checking, bank account or payment data consolidation, and supplier onboarding. These are also areas where automation and AI could ease pressure, provided controls and oversight are strong enough.
The tension between automation and control helps explain why adoption is rising while maturity lags. 40% of CFOs say they already use AI in formal finance workflows, but only 12% rate their ability to use AI effectively in core finance workflows as advanced or highly mature.
This isn’t something to put on the back burner. Nine in ten CFOs said they are under pressure from the board, CEO, or wider business to adopt AI, and more than a third describe that pressure as significant. The risk is when pressure to adopt outpaces readiness.
The most attractive use cases are practical and specific
CFOs surveyed prefer AI use cases that are focused and operational. They want to root out duplicate payments, improve error detection, streamline payment workflows, improve cash application and reconciliation, improve cash visibility and forecasting, and detect fraud or anomalies. These are specific finance pain points where automation, pattern recognition, and workflow support could produce measurable value.
The conditions for adoption are equally specific. CFOs want to trace AI outputs back to source data or underlying transactions, start with one use case before expanding, see evidence from similar organisations, measure ROI or time savings, and integrate AI with existing finance systems.
The business case is strongest when AI is tied to outcomes such as faster collections, improved cash-in, better forecasting and visibility, improved working capital, faster reporting, and scaling without adding headcount. CFOs may support AI investment most readily when it is tied to finance performance rather than vague transformation language.
Moving too quickly could create new risks
The risks of moving too quickly with AI are stark: harder-to-detect errors, decisions based on poor data, loss of confidence in the numbers, inaccurate forecasts or cash-flow decisions, increased fraud or payment risk, weak audit trails and limited explainability. These concerns go to the heart of the CFO’s role: maintaining trust in financial information.
More broadly, finance leaders see AI as increasingly necessary, but they also see the limits of adopting it before data, systems, processes, and governance are ready. Large majorities agree that fragmented systems reduce resilience, that CFOs will become more accountable for proving AI ROI, and that AI should not operate within finance workflows without human approval, audit trails, and accountability.
For CFOs, the challenge is not whether to adopt AI, but how to do so responsibly. The opportunity is real, especially in cash visibility, forecasting, reconciliation, payments, collections, and reporting. But broad automation is more likely to advance through targeted use cases, underpinned by clean data, interoperability, governance, and human accountability.
