The treasury conversation around artificial intelligence has changed.
The useful question is no longer whether AI will affect treasury. It is where it can support real work, what your team needs before it can be trusted and how responsibility remains clear.
The Association of Corporate Treasurers highlighted practical use cases in May 2026, including cash forecasting, scenario modelling, FX monitoring, liquidity digests and idle-cash detection. J.P. Morgan has also described the movement towards AI-supported corporate cash operations, while stressing that governance and accountability are still developing.
AI may help your team:
These applications can help your team interpret more information without manually reviewing every record.
The value is not simply another prediction or dashboard. It is faster understanding of what changed, why it matters and where attention may be required.
AI cannot compensate for incomplete balances, inconsistent transaction descriptions or conflicting classifications.
If the underlying information is weak, an AI model may process those weaknesses more quickly.
A practical starting point is to bring the required information together and transform it into a consistent financial structure. That may involve bank data, ERP records, invoices, forecast submissions and market information.
Context matters. A movement becomes more useful when the system can connect it to the correct entity, obligation, customer or forecast driver.
AI and connected data can support a more continuous forecasting process.
Instead of presenting only one expected number, your team may be able to explore the main drivers, confidence ranges, changes since the previous forecast and differences between expected and actual movements.
This does not guarantee perfect forecasting. Treasury rarely has complete certainty over future receipts, payments or market movements.
The benefit is a clearer understanding of what is likely to happen, why the position may be changing and where uncertainty is increasing.
AI should not be treated as an independent decision-maker.
Your team still needs to understand:
The UK Treasury Committee’s January 2026 report on AI in financial services highlighted the opportunities and risks of adoption, including the need for clearer oversight and resilience. Its April 2026 follow-up included responses from HM Treasury, the FCA and the Bank of England.
For treasury, AI may analyse information, surface risks and present possible courses of action. People responsible for cash and liquidity should retain control over material decisions and execution.
Start with a defined use case where information is available, the output can be checked and the benefit is measurable.
Forecast-versus-actual analysis, liquidity monitoring or exception identification may provide more practical starting points than attempting broad autonomy.
You should also define what success means, how outputs will be reviewed and what happens when the system is uncertain or wrong.
Fennech Treasury Intelligence helps your team understand where cash is held, how it is moving and which areas may require attention.
The Fennech Forecasting Agent can support analysis of historical patterns and cash flow drivers, confidence ranges and forecast-versus-actual comparison. It is designed to support professional judgement, not replace it, and it does not independently execute treasury decisions.
Speak to Fennech about applying AI to a defined treasury problem while keeping control and accountability with your team.
