AI for treasury management: Why maturity matters for ROI

Why mature treasury operations matter for AI value

As companies increase in treasury complexity by adding more transaction volume, legal entities, systems and banking relationships, some of their processes can be slow to keep up. Finance leaders are increasingly trying to close that gap using AI for treasury management, but is the treasury function mature enough to act on those insights?

Three themes are emerging from our analysis and conversations with clients across industries: AI value depends on treasury maturity, systems integration turns data into action, and liquidity management preserves a company’s ability to pivot and respond.

Market insights and client perspectives on AI

To contribute on-the-ground experiences to our analysis, we invited two First Citizens Bank clients to share their perspectives and recent examples from their work. Jerry Caruso is VP, Global Controller at Bitsight, a global leader in cyber risk intelligence. Ran Rabinowitch is Head of Finance for the device business at Wyze, an innovator in consumer electronics.

The promise of AI for treasury management is clear. It enables teams to better understand and act on patterns, as well as uncover issues that require deeper attention. For many companies, AI has become a strategic imperative. Our "State of the VC-Backed CFO" report noted that among VC-backed companies, "62% ranked AI adoption as a top-two issue, up from 49% from last year." The pattern is similar across other industry verticals. KPMG's 2026 Global AI in Finance study looked at enterprises across 13 industries (such as manufacturing, healthcare and energy) and found that 75% are using AI for finance functions.

The finance tech stack is evolving beyond systems that primarily record and report information into systems that enable teams to initiate actions. And this becomes more crucial as a company grows. It also raises new questions about controls, authorization, auditability and the role of human oversight as AI gets deployed.

5 capabilities to make treasury AI-ready

In practice, capturing value from AI comes down to a set of core capabilities for finance teams.

1. Visibility – to maintain a clear, real-time view of cash position. An integrated view of cash management enables companies to track deposits, liquidity, loans, credit and payables in one place. This represents the ground truth to support decisions and take action.
2. Connectivity – to seamlessly move money and data across banks and internal systems. Treasury API integration and other connectors help teams automate processes to reduce manual work and improve data quality.
3. Control – to govern finance operations, risk management and compliance. Managing treasury at scale requires controlled oversight over user access, approvals, exceptions and more.
4. Optimization – to better manage capital, liquidity and payments. It may include improving how a company structures systems, processes and data flows to manage and deploy operating and idle cash.
5. Intelligence – to unlock business value from data. With unified financial data, organizations gain insights to make more timely, informed decisions for liquidity planning, forecasting and risk.

As a company grows, keeping a clear view of cash gets harder. Payment volumes increase, more accounts must be monitored, and cash balances may become larger and perhaps more volatile. That puts pressure on treasury teams to understand where cash is and what they may need next. Manual reporting and reconciliation processes often spring up in response, which have the long-term impact of undermining visibility. Successful teams focus on integrated systems and centralized data to see cash across the business and make better liquidity decisions.

AI can surface useful insights, but acting on them is harder when information and workflows are fragmented.

Our internal First Citizens Bank analysis shows that API adoption increases among clients with greater treasury and operating complexity. API usage reaches 14.3% among middle-market companies and nearly 21% among larger companies operating globally. The pattern suggests that API use is associated with treasury maturity and that integrated connectivity becomes increasingly important as transaction volumes, systems, and governance requirements grow.

Automating workflows

Companies use AI to automate routine tasks and treasury workflows such as reconciling transactions and invoices, identifying exceptions and routing them to the right person. KPMG's Digitization in Accounting study found that among businesses using AI to automate accounting work, 70% reported improved process quality and 37% noted significant time savings after implementation.

Connecting a client's bank, ERP and treasury systems through APIs and other integrations can improve the quality, accessibility and consistency of data AI can use. Connectivity in turn supports automation across finance functions including receivables, payables, reporting and reconciliation.

Ran Rabinowitch described how his team is building internal solutions to reduce manual work. After the monthly close, there used to be a series of tasks to generate reports.

"Now, clicking the close button can send an agent to pull and update tables using predefined queries, save them to a predefined template, and then generate the report."

The impact is significant.

As work is automated, controls around access, approvals, fraud and exception handling become more important. In a 2026 Deloitte study, 59% of CFOs said their biggest AI governance challenge is balancing adoption with risk.

Jerry Caruso offered a concrete example of the controls that become important. Data integration has to be managed with appropriate permissions.

Treasury controls must also define when a human review is needed. Decisions involving payment authority, account structures, policy changes or material exceptions are good examples.

Managing liquidity and capital well can be critical to sustaining growth. The journey to profitability takes time, and financial execution has to demonstrate the discipline to support it. Among firms earning $10 million to $25 million in revenue, our latest CFO report found that 74% are still in the red, and 20% have a 1-year gap in the runway needed to reach profitability. Bridging that gap, or avoiding it entirely, can depend on how effectively a company manages cash and deploys capital.

As companies grow, finance leaders may need to rethink where cash sits and how efficiently money moves.

However, analytics and liquidity planning are challenging use cases for AI in treasury management. Cash flow patterns fluctuate shaped by stages of growth, business models, supply chain strategies, expansion goals and other factors. We continue to think this is an area that benefits from human judgement and collaboration.

Improving AI cash flow forecasting and analysis is possible if the underlying financial data is reliable. But companies are still struggling with ROI.

AI ROI may show up as greater analytical depth, faster cycles and more work completed with the same team.

Traditional productivity metrics may not fully capture all of these benefits.

CFOs may need a tiered ROI framework: lightweight evidence for experimentation, but increasingly rigorous business cases as AI spending or deployment scales. Ran described a more explicit AI investment model in which greater spending requires greater proof.

In our recent CFO report, we found that even among more mature companies, 63% still rely heavily on spreadsheets for budget planning. This is consistent with findings from the Association of Financial Professionals report that notes that across industries, 96% of FP&A professionals use spreadsheets as a planning tool at least weekly. Spreadsheets often make up for gaps in data flows and integration.

When data is better curated and fed by real-time access, AI and treasury analytics can help treasury teams see patterns earlier, improve forecasts and make more informed decisions on liquidity, capital and risk. This also better positions them to respond to the increasing urgency to implement AI solutions.

Balancing AI with human expertise

Using AI responsibly includes a clear awareness that accountability remains with finance leaders.

Where does human experience matter most? It concentrates in critical decisions such as setting policies for liquidity, investments and risk. Expertise is also essential in high-stakes moments, like resolving material exceptions related to an enterprise customer, dealing with a service disruption, or managing a data breach crisis.

Combining human judgment and AI intelligence brings together the best of both worlds.

AI in treasury—ROI depends on more than tech

AI adoption in finance is accelerating, but the economic results remain uneven and hard to track. That gap is important for treasury leaders who are usually asked to turn vast amounts of data into precise forecasts and automated recommendations for better liquidity decisions. The tech may be new, but getting value still depends on operational fundamentals.

Building tangible value through workflow automation, better forecasting and working capital management requires a strong foundation. To turn AI intelligence into informed action, companies need mature, scalable treasury operations that combine integrated solutions, reliable data and strategic expertise.

Combining treasury automation and AI intelligence with expert human judgment helps you capture more ROI as complexity increases. AI can identify and propose a response, but leaders need to decide whether and how to act. This combination provides greater agility to sustain growth, solve problems quickly and act on opportunities amid challenging markets.