Treasury has long been one of banking’s most critical but least visible functions. It manages liquidity, funding, cash positioning, risk, settlement, and balance-sheet discipline. Now, as real-time payments, volatile markets, digital assets, and AI-driven decision-making reshape financial services, treasury is moving from a back-office control function into a strategic technology battleground. The reason is simple: banks can digitise customer journeys and automate compliance, but if liquidity, cash forecasting, and settlement operations remain slow, fragmented, and spreadsheet-heavy, the whole institution remains constrained.
Treasury Is Becoming A Real-Time Function
Traditional treasury operations were built around periodic reporting, end-of-day balances, batch settlement, and scheduled liquidity reviews. That model is increasingly out of step with modern banking.
Instant payments, 24/7 transaction flows, shorter settlement cycles, and digital asset activity are forcing treasury teams to operate continuously rather than periodically. Industry commentary on treasury transformation now points to a shift from end-of-day reporting toward intraday and continuous cash visibility across banks, currencies, entities, and geographies.
For banks, this creates a major operational challenge. Liquidity positions can change faster than legacy treasury systems were designed to track. Payment flows can move outside traditional banking hours. Corporate clients expect real-time visibility. Regulators expect stronger control. Senior management expects better capital efficiency.
This is where AI starts to become strategically important.
AI Could Change Liquidity Forecasting
Cash and liquidity forecasting remains one of the hardest problems in treasury. It depends on multiple data sources, including payments, deposits, market movements, customer behaviour, loan activity, funding needs, corporate flows, and macroeconomic conditions.
Historically, much of this work has relied on manual modelling, spreadsheets, treasury management systems, and analyst judgement. AI offers the potential to improve forecasting by detecting patterns across larger datasets, identifying anomalies earlier, and producing more dynamic liquidity scenarios.
J.P. Morgan has argued that AI is bringing greater precision and efficiency to cash and liquidity management, particularly around forecasting and planning. PwC’s 2025 Global Treasury Survey also found that treasury teams are adopting AI-enhanced forecasting, real-time liquidity tools, in-house banking models, and centralised payments to improve cash efficiency and unlock trapped cash.
For banks, better forecasting is not just an efficiency gain. It can influence funding costs, capital deployment, intraday liquidity buffers, risk appetite, and client servicing.
Corporate Clients Are Raising The Bar
The pressure is not only internal. Corporate clients are also changing what they expect from banks.
Treasury teams inside large companies increasingly want real-time visibility, stronger cash forecasting, better liquidity management, and banking services that integrate with ERP and market data systems. Mastercard has noted that real-time visibility is no longer optional for corporate treasury teams, especially as companies respond to supply-chain disruption, policy changes, and supplier payment shifts.
That creates a competitive opening for banks.
The banks that can provide AI-enabled treasury insights, predictive liquidity tools, automated cash positioning, and smarter payment orchestration may become more valuable to corporate clients. Those that cannot may risk being reduced to basic transaction providers while fintechs and treasury platforms own the intelligence layer.
Treasury Is Where AI Meets The Balance Sheet
Many AI use cases in banking focus on productivity. Treasury is different because it sits directly against the balance sheet.
AI in treasury could support:
- intraday liquidity monitoring
- cash-flow forecasting
- funding optimisation
- payment routing
- settlement risk detection
- FX exposure analysis
- trapped cash identification
- working capital insights
- stress scenario modelling
- anomaly detection across payment flows
This makes treasury one of the most commercially important areas for AI deployment. A better chatbot may improve service. A better treasury AI model could improve liquidity efficiency, reduce funding costs, strengthen resilience, and protect margins.
That is why treasury could become a serious battleground between banks, fintechs, core technology providers, treasury management platforms, and AI infrastructure firms.
The Risk Is Losing Control Of The Intelligence Layer
Banks already face competition from specialist treasury technology vendors that are embedding AI into cash forecasting, payments, and liquidity management workflows. The danger for banks is that corporate clients begin relying on external platforms for decision intelligence while using banks only for account infrastructure and payment execution.
This is similar to what has already happened in parts of payments and embedded finance. The institution holding the account does not always control the customer experience or the data layer.
If treasury intelligence becomes externalised, banks could lose influence over:
- corporate cash visibility
- payment decisioning
- liquidity optimisation
- FX timing
- working capital advisory
- cross-border settlement strategy
That would weaken one of the most valuable corporate banking relationships banks still control.
Governance Will Decide How Far AI Can Go
Treasury is too sensitive for uncontrolled AI deployment. Banks cannot allow black-box models to move liquidity, trigger funding decisions, or recommend balance-sheet actions without governance, explainability, and human oversight.
This means adoption will likely begin with decision support before moving toward supervised automation.
Accenture’s 2026 banking trends highlight the growing role of AI agents across banking operations, but also stress the need for defined identity, access controls, and governance as agents become embedded in enterprise technology stacks.
For treasury, this governance layer will be critical. Banks will need to know what data an AI system accessed, why a recommendation was made, who approved the action, and how the decision fits within liquidity risk frameworks.
The Next Treasury Model Will Be Predictive
The future treasury function will not simply report what happened. It will increasingly predict what is likely to happen next.
That means moving from:
- static cash reports to live liquidity dashboards
- manual forecasts to AI-enhanced projections
- reactive funding decisions to scenario-based planning
- fragmented payment oversight to centralised orchestration
- periodic risk checks to continuous monitoring
PwC’s treasury research argues that the function is increasingly moving from transactional custodian to strategic value architect, supported by digital capabilities, risk management, and cross-functional collaboration.
For banks, this shift could redefine treasury’s role inside the institution. It may no longer sit quietly behind the business. It could become one of the most important control towers in modern banking.
What this means for the industry
- Treasury is becoming a continuous, real-time banking function
- AI could significantly improve liquidity forecasting, cash visibility, and funding decisions
- Corporate clients increasingly expect predictive treasury services from their banks
- Banks risk losing the treasury intelligence layer to fintechs and specialist platforms
- Governance, explainability, and human oversight will be essential for AI adoption
- Treasury could become one of the most commercially valuable AI battlegrounds in banking
Image Source: Pexels.com
