The role of treasury has undergone a profound transformation. Once viewed primarily as a back-office function focused on liquidity and risk control, treasury is now firmly positioned at the strategic core of the enterprise. Today’s treasurers are expected to not only safeguard cash, but also forecast future scenarios, guide decision-making, and help organizations navigate uncertainty in real time.
Treasury teams are being asked to operate faster and within new constraints that their cash forecasting and related processes weren’t originally designed to support. Today, a number of factors make forecasting more difficult, while at the same time, liquidity decisions are happening more frequently.
Teams are updating views of cash weekly or daily, sometimes more often, even as the underlying data remains fragmented across systems and business units. It creates a gap where treasury teams are still responsible for producing a clear, usable view of cash, but the inputs behind that view are inconsistent, distributed, and constantly changing.
This is the environment AI is entering. And it comes at a perfect time. It’s not a replacement for treasury teams, but will instead help teams enhance their current processes to meet these new demands. With AI acting as a support layer, teams still remain in control over operations and decision-making.
What Defines a Modern Treasury Team?
Modern treasury teams are defined less by transactional processes and more by their ability to generate insight. Industry research consistently points to this shift: treasury is evolving into a strategic partner that contributes directly to enterprise value creation, driven by data, connectivity, and advanced analytics.
For the modern treasury team, forecasting is no longer a single output. It has now become a set of overlapping views where short-term liquidity forecasts, medium-term projections, and long-term planning assumptions all exist at the same time. Each is built differently, with different levels of precision and different inputs.
So the challenge then becomes making sense of them together, where long-term forecasts may be accurate, but are compared against shorter-term forecasts that are less precise.
Treasury sits at the center of these forecasts. That means working across finance, operations, tax, and commercial teams to build a view of cash that is good enough to act on. But it also means treasury faces greater structural complexity, as cash moves through different systems at different speeds and is ultimately recorded in different ways, creating new layers of challenge.
Even in well-managed environments, there is rarely a perfect match between inflows and outflows. This is likely why 62% of treasury teams reported cash and liquidity forecasting as their most pressing challenge, and an additional 46% of teams are evaluating AI to help with cash forecasting.
Other areas that treasury would like to see AI support include:
- Process automation
- Information summarization
- Improving data quality and decision-making
- Improving payment processes
Despite the shift toward strategy, the reality of most treasury teams still involves a heavy operational workload. Many organizations continue to rely on fragmented systems, manual workflows, and spreadsheet-driven processes to manage essential tasks such as cash positioning and forecasting.
Many forecasting workflows still rely on manual data collection and reconciliation, where inputs are pulled from different systems before they are adjusted and aligned for use. Each step introduces delay and variability. So the primary constraint on forecasting accuracy is not modeling capability; it is instead data quality, consistency, and integration.
How Treasury Teams Are Actually Using AI
In practice, treasury teams are not using AI to replace forecasting, nor do they want to. Instead, teams use AI to make the process more manageable, and the most immediate impact will likely be for short-term forecasting. High-frequency, data-heavy use cases are where AI performs best.
Treasury can use AI to identify patterns in collections and disbursements more quickly. Or AI can be used to help update forecasts more frequently without rebuilding assumptions from scratch. Even variances can be flagged earlier.
AI has a role in longer-term forecasting, but it looks different from short-term use cases. The value of AI for long-term forecasts is less about precision and more about adding context. By analyzing historical patterns and broader trends, AI can help treasury teams test assumptions, model different scenarios, and understand how changes in the business or market environment might impact future cash positions. AI can make it easier to compare outcomes and pressure-test decisions.
AI as Support Layer: Why AI Doesn’t Reduce Treasury’s Role
Though AI offers significant value to modern treasury teams, only 26% of teams consider their AI capabilities to be moderately or very mature. It also makes existing issues harder to ignore.
AI doesn’t resolve fragmented data environments or inconsistent assumptions across the business. Unfortunately, it can really amplify them. Poor inputs lead to unreliable outputs, only faster and on a greater scale. This reality is one of the reasons adoption has been measured.
When teams can’t trust their data quality, then they can’t trust in AI-generated outputs. This is where the role of treasury becomes more, not less, important. While AI can change how forecasts are built and automate some data collection processes, it can’t change who is responsible for those forecasts and the data.
While AI can help teams move faster and see more, there are still some non-negotiable functions that need to be owned by the humans on the treasury teams:
- Assumptions behind the forecasts
- Defining what is being measured
- How inputs are gathered
- How results are interpreted
- Deciding how to act when forecasts change
- Data quality and consistency
As data environments improve and integration challenges are addressed, AI will continue to expand its role within the forecasting process. In the near term, it will remain complementary, supporting existing workflows, improving efficiency, and creating greater transparency.
Finding the Right Balance
AI makes it easier to see where assumptions diverge, where data is inconsistent, and where forecasts may not reflect underlying cash flow behavior. That visibility brings both opportunity and pressure, raising expectations around accuracy, speed, and discipline.
If your forecasting process is being stretched by fragmented data, increasing cadence, and higher expectations, AI can help. The value is not in replacing treasury, but in giving teams a clearer, faster way to understand and act on their data.
TIS helps treasury teams connect systems, improve data quality, and apply AI in a way that supports real decision-making. Book a demo to see how it works in practice.
