Treasury Management Systems in Transition Through Artificial Intelligence

A clear technological development path is emerging in corporate treasury: treasury management systems (TMS) are gradually extending their traditionally function‑ and process‑oriented user interfaces with AI‑enabled interaction models. The aim of this development is not to replace existing systems, but rather to change how they are used – moving away from manual navigation across individual functions toward an information‑, event‑ and decision‑centric interaction model that actively supports treasurers in analysis, prioritization and decision‑making.

In this article, we examine the ongoing evolution of treasury management systems (TMS) and highlight key success factors shaping this transformation.

It is hardly surprising that artificial intelligence is at the center of TMS development. AI is increasingly seen as indispensable and is being designed as an integral component of system architectures.

Broadly speaking, three complementary layers of AI deployment can be observed: 

1. Embedded analytical AI for pattern recognition and anomaly detection,
2. Generative assistant functions that translate complex topics into clear, contextualized information as well as
3. AI agents that autonomously execute preparatory or rule‑based tasks. 

These layers are increasingly shaping the user experience of modern TMS – for example in ERP‑adjacent platforms such as SAP with its AI assistant Joule, or in specialized treasury solutions such as Kyriba with agent‑based AI.

At the same time, the underlying logic of TMS is not being fundamentally questioned. The foundation continues to be an ecosystem integrated via interfaces, comprising finance functions, transactional modules, market data, trading platforms and reporting solutions. The additional AI layers described above, however, fundamentally change how users interact with these systems.

Regardless of the provider, a consistent trend is emerging: future user interfaces will shift the focus away from manual system operation toward dialog‑driven, event‑ and decision‑centric interaction. Treasurers are supported by AI‑driven insights, recommendations and prepared actions, while retaining full professional responsibility.

The following sections outline selected ideas and impulses on how these developments may affect concrete treasury processes and the opportunities that may arise.

Bank Account Management
In bank account management, future user interfaces primarily enhance transparency, control and compliance. AI‑enabled interfaces consolidate account information across banks and countries, identify discrepancies in master data or signing authorities, and proactively flag overdue confirmations or reviews. Requests to open or close bank accounts can be prepared through dialog‑based workflows, while approval and control responsibilities remain with the treasurer.

Cash Management and Liquidity Steering
In cash management, the main value lies in the consolidation of complex information. AI‑based user interfaces analyze cash positions, forecasts and cash flows in real time, identify potential liquidity gaps or surpluses, and present them in a clear context of causes and potential actions. Treasurers receive indications of possible measures, such as internal fund reallocations or short‑term financing, without triggering automatic execution. This is particularly relevant for external cash management, as many intra‑group processes are already largely automated.

Payments
In payments, the focus shifts from manually monitoring individual transactions to event‑driven control. AI‑enabled interfaces detect irregularities, potential errors or unusual patterns at an early stage and prioritize them for user review. In exception cases – such as rejected or unusually large payments – causes are analyzed and clearly explained, enabling targeted intervention.

Financial Risk Management
In financial risk management, the TMS supports the analysis of FX, interest rate or commodity risks by consolidating exposures, market data and hedging instruments. AI can also interpret unstructured data and present it in a comprehensible manner. Scenarios can be queried through dialog‑based interaction, for example to assess the impact of changing market conditions. Recommendations on hedging strategies remain transparent and explainable, with final decisions clearly retained by the treasurer.

Debt & Investment Management
For financing and investments, a consolidated view of maturities, terms, covenants and risks can be provided. Deviations, upcoming maturities or covenant risks are highlighted at an early stage with AI support. The primary value lies less in automated decision‑making and more in the structured preparation and prioritization of relevant information.

Internal Controls and Compliance
In internal controls and compliance, AI‑enabled user interfaces support continuous monitoring of controls, authorizations and rule breaches. Deviations are prioritized based on risk and supplemented with explanatory guidance, improving transparency without introducing additional manual control steps. Looking ahead, AI could also support automated alignment with current and future regulatory requirements, flagging potential violations and suggesting mitigation measures.

The benefits of AI‑enabled user interfaces depend on several prerequisites, including high data quality, consistent master data, integrated system landscapes, and clearly defined roles and authorization concepts. Transparency is also critical: the logic behind analyses, alerts and recommendations must be explainable in order to build trust and acceptance.

At the same time, value must be delivered quickly and complex corporate data structures must be effectively managed. These capabilities need to be provided natively within modern, cloud‑based treasury systems, reducing the traditional effort required for upgrades and rollouts. Regular releases ensure that new features are continuously deployed in live operation. 

The integration of artificial intelligence into modern treasury management systems primarily changes how treasurers interact with their systems. Where complex navigation and detailed system knowledge were previously required, AI‑enabled interfaces now allow for direct, dialog‑based interaction. Professional questions can be asked in natural language, relevant data is aggregated, relationships are contextualized, and results are presented clearly. New employees in particular benefit from this interaction model, as guided workflows, intuitive queries and automatically generated summaries significantly shorten onboarding into treasury processes.

• Direct access to treasury‑relevant information without system navigation
• Productive use without deep technical expertise
• Context‑based workflows and action suggestions instead of isolated system functions
• Shorter onboarding periods through guided interaction
• Structured, easy‑to‑understand summaries to support decision‑making

Treasury management systems are evolving from transaction‑focused applications into interaction‑ and decision‑centric user interfaces. Treasurers increasingly work with an intelligent interface that structures information, prepares action options and supports decisions. High data quality remains a critical prerequisite for sustainably leveraging the potential of AI. From our perspective, AI can complement a fully integrated banking and system landscape – but it cannot replace it at this stage.