| DAY 1 — The Transformation: What Big Data and AI Make Newly Possible |
| 8:30 – 9:00 | Registration and Light Breakfast |
| 9:00 – 9:10 | Welcoming Remarks — Senior IMF Management |
| 9:10 – 9:20 | Introduction to the Forum — Director, IMF Statistics Department |
| 9:20 – 10:15 | Session I: Opening Keynote — The Measurement Inflection Point: Economic Statistics in the Era of Big Data and AI |
| | Economic measurement is at an inflection point. Big data and artificial intelligence are not simply improving existing statistical processes at the margin; they are expanding the frontier of what can be known about an economy, how quickly it can be known, and at what level of detail. New data sources, advanced computational methods, and AI-enabled tools are creating possibilities for more timely, granular, and policy-relevant insights into economic activity. The discussion will challenge participants to move beyond the question of how to make current statistical systems more efficient, and instead ask what new forms of measurement, analysis, and policy understanding are now possible. It will provide a common foundation for the Forum, linking innovation in data production to broader questions of policy relevance, institutional readiness, trust, and the future role of official statistics. |
| 10:15 – 10:45 | Coffee Break |
| 10:45 – 12:15 | Session II: New Data, New Policy Options: Statistics for Fiscal, Financial, and Monetary Policy |
| | Big data and artificial intelligence are expanding the information available to fiscal, financial, and monetary policymakers, creating opportunities to develop more timely, granular, and policy-relevant datasets. This session will examine how countries are using new data sources and AI-enabled methods to address current fiscal, financial and monetary policy needs and persistent data gaps. Examples may include real-time monitoring of revenues and expenditures, analysis of tax compliance gaps and the informal economy, assessment of subnational fiscal risks, tracking of public investment, measurement of financial inclusion, and identification of emerging financial vulnerabilities. The session will also consider how scanner data, web-scraped prices, electronic payments, credit registries, and other granular sources can improve inflation measurement, track price dispersion and pass-through, monitor demand conditions, and provide earlier signals of financial stress and monetary policy transmission. A central theme will be how better data can enhance the policy toolkits and allow policymakers to design more targeted interventions, and act with greater confidence. |
| 12:15 – 1:30 | Lunch Break |
| 1:30 – 2:45 | Session III: New Data, New Policy Options : Statistics for Macroeconomic, Trade, and Industrial Policy in the Era of Big Data and AI |
| | Big data and artificial intelligence are creating new opportunities to measure economic activity, trade flows, production, and structural change with greater timeliness, detail, and geographic precision. This session will examine how countries are using alternative data sources and AI-enabled methods to address emerging macroeconomic, trade, and industrial policy needs. Examples may include the use of satellite imagery, mobility data, night-light data, web-scraped prices, customs records, firm-level data, shipping and logistics data, and other high-frequency sources to track GDP and its components, monitor supply-chain disruptions, assess regional and sectoral developments, measure digital and platform-based activity, understand trade dependencies, and evaluate industrial policy outcomes. A central focus will be how these new data sources and methods allow policymakers to see developments earlier, identify vulnerabilities more precisely, and answer policy questions that traditional statistics cannot fully address. |
| 3:00 – 3:30 | Coffee Break |
| 3:30 – 4:30 | Session IV: Rethinking Policy Frameworks |
| | Big data and artificial intelligence are not only changing how economic information is produced; they are also changing how policymakers frame problems, assess options, and make decisions. This panel will bring together policymakers to discuss how access to more new , timely, granular, and diverse information is reshaping the policy toolkit. The discussion will focus on how new data sources and AI-enabled analysis can help policymakers move beyond traditional policy frameworks, identify risks and opportunities earlier, design more targeted interventions, and better understand the distributional, sectoral, geographic, and firm-level effects of policy choices. |
| 4:30 – 5:15 | Cocktails / Networking |
| |
| DAY 2 — The Responsibilities: Navigating Quality, Trust, and Governance |
| 9:00 – 9:10 | Opening Remarks — Day 2 |
| 9:10 – 10:30 | Session V: Can We Trust the Numbers? Data Quality, Transparency, and Confidence in the Age of Big Data and AI |
| | As big data and artificial intelligence become more central to economic measurement, the credibility of statistical outputs will depend on whether policymakers, markets, and the public can understand and trust how those outputs are produced. This panel will examine what data quality means when statistics are derived from non-traditional data sources, algorithmic methods, and models that may be difficult to explain, reproduce, or revise. Panelists will consider how big data-derived statistics compare with traditional measures, when differences matter for policy and markets, and how institutions should communicate data provenance, methodology, uncertainty, limitations, and revisions. The discussion will also explore how data adequacy frameworks may need to evolve in a big data environment, including how institutions such as the IMF should assess quality, manage model risk, and address the challenges of real-time statistics in a world of market sensitivity and algorithmic trading. |
| 10:30 – 11:00 | Coffee Break |
| 11:00 – 12:30 | Session VI: Who Produces Statistics? The Expanding Data Ecosystem |
| | The production of economically relevant statistics is no longer confined to traditional official statistical agencies. Fintech firms, digital platforms, data aggregators, research institutions, think tanks, international organizations, public-private partnerships, and other non-official or quasi-official producers are increasingly generating data used to understand economic activity, financial conditions, prices, mobility, trade, labor markets, and household and firm behavior. This session will examine the role of these producers in the evolving statistical ecosystem, with a particular focus on how they establish quality, transparency, methodological rigor, continuity, and responsible data governance. Contributions will explore how data are sourced, processed, validated, documented, and made accessible, and how users can assess reliability, bias, coverage, confidentiality protections, and fitness for policy or analytical use. A central question will be how trust can be built and maintained when economically important statistics are produced outside the traditional boundaries of official statistics. |
| 12:30 – 1:45 | Lunch Break |
| 1:45 – 2:45 | Session VII: From Pilots to Production — Integrating Big Data and AI into Statistical Production |
| | This session explores how national statistical agencies, central banks, and other official data producers are moving from experimentation with big data and AI to their systematic use in regular statistical production. Contributions may examine how these tools are being integrated into core workflows, data architectures, quality assurance processes, and organizational arrangements. Topics may include the use of scanner or transaction data in price index compilation, machine learning for classification and editing, AI-assisted processing of administrative or financial data, and the use of geospatial, text, image, or satellite data at scale. The session is especially interested in practical lessons on governance, methodology, infrastructure, skills, risk management, and institutional models that can support the responsible adoption of these methods. |
| 2:45 – 3:45 | Session VIII: Measuring AI and Big Data |
| | This session turns the lens around: rather than using AI and big data to improve measurement, it examines how official statistics can capture AI and data themselves as economic phenomena. As AI moves from frontier technology to general-purpose infrastructure, and as data becomes a core input into production, statistical systems face new conceptual and practical challenges in measuring their scale, diffusion, and value. Contributions may examine approaches to measuring AI production and the AI-producing sector, the adoption and diffusion of AI across industries and firms, and AI-related investment, including intangible and complementary investments in skills, software, and organizational capital. The session also welcomes work on the valuation of data—both as an economic asset in national accounts and as a critical input into the training of large language models and other AI systems—as well as on the boundary issues this raises for GDP, productivity, and capital measurement. Topics of interest include survey instruments and questions to capture AI use, the use of administrative and commercial data sources, classification challenges (e.g., what counts as "AI"), cross-country comparability, and methodological frameworks for valuing data stocks and flows. The session is especially interested in practical experiences, pilot results, and emerging conceptual frameworks that can help official statistics keep pace with the rapid transformation of the digital and AI economy. |
| 3:45 – 4:00 | Coffee Break |
| 4:00 – 5:00 | Session IX: One on One Discussion — Embracing transformation responsibly The closing session will examine how big data and AI are transforming what can be measured about the global economy, with particular attention to the urgent need for new data on AI adoption, diffusion, and use across countries, sectors, firms, and workers. The discussion will consider what the global statistical system must do to embrace this transformation responsibly, including how to foster collaboration across institutions, integrate new data sources and methods, maintain quality and trust, protect confidentiality, and ensure that innovation strengthens rather than undermines the credibility of official statistics. |
| 5:00 – 5:15 | Closing Remarks |
