Precision Over Speed: Why Treasury Leaders Are Reframing AI Around Trust

Treasury technology has been traditionally been judged by speed: faster payments, quicker reconciliations, and achieving visibility in real time. But as risk exposure increases and global liquidity environments grow complex, CFOs and treasury leaders are recalibrating their priorities. Accuracy, explainability, and trust in data are now just as critical as velocity.

This shift reflects treasury’s expanding role within the enterprise. Once viewed as a back-office execution function, treasury now plays a central role in liquidity strategy, fraud mitigation, and financial resilience. As AI becomes more prevalent, treasury leaders are increasingly focused on how intelligence is applied, not just how quickly outcomes are delivered.

In a recent SAPinsider’s Expert Exchange, Bob Stark, Global Head of Market Strategy at Kyriba, and Jean-Baptiste Gaudemet, Senior Vice President of Data and Analytics at Kyriba, spoke with Susan Galbraith, Vice President and Research Analyst at SAPinsider, about how treasury teams are using AI to improve confidence in decision-making rather than simply accelerating transactions.

Trust Is the Gatekeeper for AI Adoption

Stark emphasized that AI initiatives often stall because of human skepticism, not technical limitations. “If treasury teams don’t trust the data behind an AI recommendation, they won’t act on it,” he said. “And if they don’t act on it, the automation simply doesn’t matter.” From Kyriba’s perspective, this places governance and explainability at the center of any treasury AI strategy.

Gaudemetreinforced this point from a data architecture standpoint. “Treasury data doesn’t live in isolation,” he explained. “Cash positions, bank connectivity, payment behavior, and historical trends all inform one another. AI has to understand those relationships, not just individual data points, to produce insight teams can rely on.”

As payment volumes increase and networks grow more complex, this contextual understanding becomes essential. Without it, treasury teams risk reacting to noise rather than identifying genuine risk or opportunity.

From Anomaly Detection to Pattern Recognition

The conversation also explored how AI is reshaping payments intelligence. Gaudemet noted that traditional systems often overwhelm teams with alerts. “The real value comes from understanding what ‘normal’ looks like for your organization,” he said. “Once you establish that baseline, meaningful exceptions become obvious.”

Both Kyriba experts stressed that technology cannot compensate for fragmented data environments. Clean master data, consistent definitions, and disciplined processes remain prerequisites. As Stark put it, “AI amplifies whatever environment you put it in—good or bad.”

The discussion also pointed to a shift on the horizon in treasury skill sets. As automation absorbs routine execution, treasury professionals will increase their focus on interpretation, exception management, and strategic guidance. In this model, AI doesn’t replace human judgment: It sharpens it.

What This Means for SAPinsiders

AI readiness in treasury is a data governance question first, not a tooling decision. The conversation underscored that AI  adoption in treasury consistently stalls when underlying data cannot be trusted. As both Kyriba executives noted, treasury data spans cash positions, bank connectivity, historical payment behavior, and regional variations, which makes consistency critical. SAP finance leaders should resist the temptation to pilot AI on fragmented datasets, where inconsistent definitions across entities or geographies undermine confidence in outputs. Before introducing advanced analytics or AI-driven recommendations, organizations must standardize data models, reconcile sources, and establish governance frameworks that ensure explainability. Otherwise, applying AI becomes an academic exercise that treasury teams quietly sidestep rather than rely on for decisions.

Payments intelligence must evolve from alerting to interpretation. Real-time visibility alone does not improve treasury performance if it just generates more alerts. As discussed in the exchange, the real challenge for treasury teams is to distinguish meaningful exceptions from background noise. SAPinsiders should prioritize architectures that establish behavioral baselines. These include discerning what “normal” cash movement looks like for their organization across banks, geographies, currencies, and counterparties. This contextual understanding allows AI to highlight genuine risk rather than surface every deviation. The shift to interpretive intelligence over alert-driven monitoring enables treasury teams to move from reactive firefighting toward proactive liquidity oversight and smarter decision-making.

Treasury talent strategy must evolve alongside automation. As AI absorbs execution-heavy tasks such as reconciliation and monitoring, treasury professionals are empowered to shift toward interpretation, risk assessment, and strategic advisory work. Kyriba’s experts emphasized that AI does not eliminate human judgment: It increases its importance. SAP leaders should plan for reskilling that emphasizes analytical reasoning, cross-functional collaboration, and the ability to evaluate AI-driven insights in business context. Organizations that treat automation as a substitute for expertise risk hollowing out their treasury capability. But those that invest in talent development as well as technology position treasury as a strategic partner in enterprise resilience, liquidity planning and growing profitability.