Credit iStock.com/metamorwoks
Credit iStock.com/metamorwoks
Credit iStock.com/metamorwoks
Credit iStock.com/metamorwoks
Digital treasury reform is not just about new systems or smarter dashboards but starts with strong data foundations. That was the key message of the Third ASEAN Treasury Forum (ATF) in July 2026, which brought together treasury officials from across the region to explore how clear data ownership, common standards, reliable information flows, and practical use cases for analytics and artificial intelligence can make digital public financial management work better.
Finance ministries and treasuries are under growing pressure to use data for better decision making. They need real-time visibility of cash positions, more accurate liquidity forecasts, stronger payment controls, and timely evidence to guide fiscal decisions. Emerging technologies such as machine learning and artificial intelligence can help, but only if they are built on data that are complete, sufficiently granular, consistent, accessible, and well understood. Where treasury data are fragmented across systems, classified inconsistently, reconciled manually, or difficult to compare, technology does not fix the problems; it magnifies them.
This connection between data quality, technology, and PFM outcomes was the central theme of the Third ATF, held virtually on July 16, 2026. Hosted by the Philippines Bureau of the Treasury with support from the IMF’s Fiscal Affairs Department and Capacity Development Office in Thailand, the Forum brought together treasury representatives from the 11 ASEAN countries and more than 200 participants. The discussion focused on a simple but powerful idea: stronger data foundations are key to making analytics, interoperability, and AI truly useful for modern treasury management.
The starting point was not technology for its own sake, but the public financial management problems that better data and digital tools can help solve. Ahead of the main session, countries took part in two virtual pre-clinics, held on June 15 and July 1, and completed diagnostic exercises to identify their most pressing treasury management challenges. A common set of issues emerged across the region: fragmented data ownership, weak interoperability across systems, inconsistent standards, and difficulty producing timely information for decision-making. These are not just technical issues; they are core treasury challenges that affect liquidity management, budget execution monitoring, fiscal reporting, payment controls, and risk management.
AI was discussed not as an end in itself but as a practical tool to help treasuries detect and manage risks, improve forecasts, and focus attention on issues requiring closer scrutiny. For finance ministries and treasuries, its immediate promise is not replacing human judgment but helping officials detect patterns and anomalies that warrant further verification, whether in contracts, supplier information, payment records, or cash flow trends. At the Forum, an Agent AI demonstration showed how contracts could be reviewed to detect unusual patterns and potential red flags such as contract splitting, vendor affiliation, and duplicate payments. An AI-supported cash forecasting model also illustrated how machine learning can identify patterns in historical cash inflows and outflows that traditional methods may miss. These examples underscored a clear message: AI tools can strengthen payment controls, improve cash forecasting, and support better decisions, but their value depends less on the sophistication of the model than on complete, consistent, timely, and well-governed treasury data.
This is why data governance matters. Better decisions, and better AI-supported insights, depend on data that are clearly defined, consistently maintained, and fit for purpose. Finance ministries and treasuries need to be clear about which data matter most: cash balances, commitments, payments, arrears, debt service, revenue inflows, and bank account information. For each of these data domains, they need clear ownership, common definitions, access rules, and quality controls. The Forum’s data clinic turned these principles into practice, asking country teams to identify the PFM problem their data governance framework should solve and the standards needed to make information comparable, consolidated, and usable. A practical data governance agenda must start with tomorrow’s analytical needs to inform decision making, not yesterday’s reporting habits. Immediate steps may include clarifying priority data domains, documenting data flows, and identifying owners of datasets. Medium-term reforms may focus on common classifications, interoperability arrangements, and stronger links between treasury systems and banks. Over time, these foundations can support more advanced analytics, automated controls, and AI-enabled decision making. The country work at the Forum followed this logic, with teams identifying immediate, medium-term, and longer-term priorities for follow-up within their own institutions.
The broader lesson is clear: digital transformation in public financial management does not start with technology, it starts with the information governments need to manage public resources well. Strong data foundations help treasuries to forecast cash more accurately, track fiscal execution more effectively, and give decision-makers timely evidence when it matters most. For ASEAN treasuries, investing in data governance is not a back-office task; it is a strategic priority to build public financial management systems that are more agile, transparent, and resilient.
