AI-Powered Treasury Transformation

“Cashflow forecasting is the primary concern among European treasurers.”

How is AI transforming treasury from spreadsheet-driven processes to intelligent, predictive decision-making?

Spreadsheets have long played an essential role in treasury management. They gave teams the flexibility to navigate fragmented IT landscapes, evolving reporting requirements and inconsistent data quality, often compensating for limitations in existing tools. However, as treasury operations grew in scale and complexity, reliance on Excel-based models, VBA macros and manual reconciliations introduced operational dependencies. This in turn diminished transparency and heightened control risks.

Given the prevailing economic environment, these challenges can no longer be overlooked. Tighter liquidity conditions combined with increased regulatory scrutiny are placing treasurers under considerable pressure to make timely, well-informed decisions. The 2026 EACT Treasury Survey reflects this reality, identifying cashflow forecasting as the primary concern among European treasurers. In response, artificial intelligence offers a clear opportunity to strengthen treasury operations. By improving how data is collected, structured and interpreted, AI enhances the accuracy and timeliness of insights. Its purpose is not to replace professional judgment, but to complement it. With greater predictive visibility, treasurers can move beyond historical reporting and contribute strategically alongside CFOs, boards and regulators, anchoring the function in forward-looking decision-making rather than retrospective control.

Which treasury functions can AI optimise today to improve cash visibility, forecasting and risk management?

AI already delivers tangible value, especially in functions still relying on manual processes and scattered information. For cash visibility, automated AI tools bring together and reconcile bank balances from different systems, offering a clearer and quicker look at liquidity. This improvement helps speed up decision-making, as reflected in the 2026 EACT Treasury Survey, where 63% of respondents identified real-time reporting as the most impactful treasury innovation, followed by real-time liquidity capabilities at 49%.

Better visibility also improves forecasting. Machine-learning models analyse historical cashflows to identify recurring patterns, explain forecast variances and strengthen responsiveness to shifting business conditions. Combined with market data, AI can take liquidity planning further by connecting forecasts with funding needs and exposure to FX or interest-rate changes. From a risk-management perspective, AI strengthens oversight of FX positions and supports more robust hedge monitoring and anomaly detection in transactional data. Automation also streamlines payment controls, reconciliations, and exception handling, mitigating operational risk while improving execution efficiency across the function.

What should treasurers prioritise when adopting AI to drive efficiency, control and strategic value?

Successful AI adoption starts with a clear and realistic ambition. Rather than pursuing immediate end-to-end automation, treasurers should focus on a limited set of high-value use cases where data quality, process maturity and governance are sufficiently robust to demonstrate measurable impact.

This ambition must then be translated into a structured roadmap. In environments characterised by fragmented system landscapes and multi-entity operating models, a phased approach allows teams to progressively improve data quality, harmonise processes and validate integrations without disrupting daily operations. Before committing to large-scale platforms, lighter solutions such as GenAI copilots, automation agents or low-code tools can help prototype workflows, test data lineage and automate reporting or controls with limited upfront investment.

As AI adoption expands, strong governance becomes essential. Data traceability, auditability and clear ownership all prove critical to meet regulatory and internal control requirements. Treasury teams must continue developing the necessary skills to understand, challenge and oversee AI-driven outputs, keeping human judgement and expertise at the core of decision-making.

Olwenn Camps-Leysour de Rohello and Romain Henrio — BearingPoint