Finance professionals are adopting AI and automation at increasing scale as they contend with widespread liquidity constraints, growing operational complexity and rising expectations around financial control. That’s the picture emerging from Panax’s 2025 AI and Automation Survey, which polled 200 senior finance professionals in the US, UK, Germany, France and the Middle East and Africa.

Respondents represented a wide range of industries, including energy, e-commerce, manufacturing, technology and hospitality. All worked at organisations with annual recurring revenue of at least $50m. Half of those surveyed did not have a dedicated treasurer on staff, while a quarter had one to three, and a further quarter had four or more treasury professionals. The findings reflect both the pressures and opportunities facing today’s finance teams, particularly as they turn to automation to improve visibility, reduce risk and respond to volatility.

Internal operational risks were the most common liquidity constraint or risk reported over the past year, cited by 33% of finance professionals. These risks include fraud and system breakdowns that disrupt cash availability. Regulatory constraints were a close second, mentioned by 32%, with legal requirements limiting access to liquid funds. Debt obligations, such as high levels of near-term maturities requiring substantial outlays, were noted by 27% of respondents. 

Other challenges included market or asset liquidity risk (26%), funding liquidity risk (24%), cash flow mismatches or short-term shortages (22%) and credit risk from counterparties (20%). The wide spread of issues suggests that treasury professionals are being pulled in multiple directions and are under pressure to tighten control and responsiveness across the board.

Forecast at your own risk

One area where these operational pressures are especially visible is cash flow forecasting. Despite growing uncertainty and rising expectations for real-time insights, many businesses continue to rely on manual or legacy systems for forecasting and reconciliation. According to the survey, 100% of companies use at least one method of cash flow forecasting, but only 26% make use of mixed-period forecasts that could help align short-term liquidity and long-term strategic planning. The most common approach is weekly forecasting, which is used by 50% of respondents, followed by monthly forecasts at 46%. Just 4% use daily forecasting models, indicating that short-term forecasting remains a minority practice despite potential benefits to liquidity management.

Tool usage reflects a shift toward automation, but not a complete one. A total of 70% of companies rely on some form of dedicated or semi-automated forecasting tool. Of these, 40% use a treasury management system (TMS), 30% use a designated forecasting platform, and 28% use Excel connected to an ERP or data warehouse. Only 2% update forecasts manually using Excel alone. 

Among companies with four or more treasury staff, 54% use a TMS, compared with 42% of those with one to three treasurers and 32% of companies with no dedicated treasury personnel. This suggests that larger teams with more complex cash flows are more likely to invest in automation.

AI: admired, feared… adopted?

The data also reveal a strong belief in the potential for AI to improve performance across multiple finance functions. Respondents were asked to identify the top areas where AI could enhance efficiency and productivity. The most commonly cited functions were audit and compliance (23%), expense management (21%), transaction categorisation (21%) and cash flow forecasting (21%). Other areas seen as ripe for improvement included budgeting and planning (20%), risk management (20%), reconciliation (19%) and financial reporting (18%). Even treasury management itself was mentioned by 18% of respondents, suggesting that no part of the finance function is immune from the drive to improve performance through smarter technology.

The perceived importance of transaction categorisation was notably higher among larger firms. While 34% of respondents from companies with revenue between $500m and $999m cited it as a priority, only 17% of respondents from companies with revenue between $50m and $499m did the same. The manual effort required for transaction categorisation scales quickly as companies grow, making it a key target for automation among more complex businesses.

Despite broad support for AI adoption, implementation challenges remain. The top three concerns cited by finance professionals were regulatory compliance (29%), data reliability (27%) and cost of implementation (also 27%). Concerns over data reliability are particularly relevant in the context of AI, where outputs are only as reliable as the inputs and training data. Many finance leaders recognise the value of human oversight to validate AI-generated insights before they are used in decision-making.

Cost concerns appear to reflect a tension between the long-term savings that automation can deliver and the short-term costs of replacing or integrating with legacy systems. While automation is often promoted as a way to cut operational expenses, implementation still requires significant upfront investment. For organisations with limited budgets or lean teams, this barrier can delay progress even where the potential benefits are clear.

Privacy and security were identified as concerns by 21% of respondents, followed by ethical and bias issues (20%), skill gaps in the workforce (16%) and job displacement or over-reliance on technology (13%). These findings reflect a common tension in AI adoption: while the technology can increase accuracy and speed, it also requires cultural and operational change to ensure successful deployment.

When asked about the top benefits of automating treasury operations, respondents highlighted a variety of factors. Leading the list was compliance with finance policies (32%), followed closely by optimising debt management to reduce interest costs (31%) and improving liquidity and credit risk management (29%). Other benefits cited included improving return on investment through more efficient cash management (28%), better management of FX risk (24%), and time saved on manual work related to data gathering and reporting (20%). Only 6% selected risk management relating to financial institutions, suggesting that internal operational improvements remain a more immediate priority than external exposures.

Spreadsheets, shortcuts and stress

Avoiding manual errors was seen as especially important by those still reliant on Excel. Among companies using Excel as their primary cash flow tool, 25% highlighted error avoidance as a key benefit of automation, compared with 18% of those using TMS platforms and 12% using designated forecasting tools. This suggests that firms still using spreadsheets are acutely aware of their limitations and more motivated to switch.

The report also sheds light on how companies manage accounts payable and receivable reconciliation. Specialised reconciliation software is now the most common tool, used by 45% of respondents. A further 39% rely on ERP system integration, and 26% use a TMS. Manual reconciliation processes are still in place at 23% of companies. 

Notably, despite many TMS platforms offering reconciliation features, the relatively low use of TMS for this purpose may point to functional gaps, siloed teams, or a lack of access among staff responsible for bookkeeping and reconciliation. This highlights the importance of providing relevant users across finance teams with the tools they need, rather than limiting automation capabilities to treasury or leadership functions.

From pain points to progress

Taken together, the survey results paint a clear picture of a finance function in transition. Liquidity pressures, operational complexity and regulatory challenges are driving demand for more efficient tools and processes. Internal operational risks were cited by 33% of finance professionals as a constraint on liquidity, while 32% pointed to regulatory barriers. At the same time, widespread concerns about compliance (29%), data reliability (27%) and implementation cost (27%) continue to slow AI adoption in some organisations.

Despite these challenges, 70% of companies are now using forecasting tools that incorporate some degree of automation. Among those still reliant on Excel, a quarter cited error avoidance as a primary motivation to automate. Businesses that remain dependent on manual processes are increasingly aware of their limitations and appear more willing to explore digital solutions.

Looking ahead, finance teams that invest in AI-driven systems are likely to gain an advantage in accuracy, visibility and efficiency. Respondents identified compliance with finance policies (32%), debt optimisation (31%) and improved liquidity and credit risk management (29%) as the top benefits of treasury automation. However, success will depend not only on choosing the right tools but also on integrating them effectively across teams, ensuring access for all relevant users, and maintaining human oversight of critical financial decisions. The opportunities are growing, but so are the expectations.