English / ქართული / русский /







Journal number 3 ∘ Nazira Kakulia Nodar Kiladze
The Impact of Artificial Intelligence and Machine Learning on Monetary Policy

journal N3 2025

DOI: 10.52340/ekonomisti.2025.03.09

Summary

This article explores the impact of machine learning (ML) and artificial intelligence (AI) on monetary policy worldwide, highlighting the increasing role of these tools in enhancing central bank efficiency and analytical capabilities. While traditional monetary policy tools are based on economic theory, expert judgment, and econometric models, ML and AI techniques are used to improve forecasting reliability, risk assessment, and policy communication.

Machine learning is efficient in macroeconomic forecasting and nowcasting by processing vast datasets and capturing non-linear relationships that evade classical models. For example, ML methods such as random forests have outperformed standard models in forecasting GDP growth and inflation, particularly in data-rich environments with volatile conditions. However, the success of ML varies with context. Overly complex models may overfit or underperform during regime shifts, underscoring the importance of careful validation and specific adjustments.

Financial stability analysis has also benefited from ML innovations. Advanced algorithms use credit risk modeling, early-warning systems, and anomaly detection for banking crises. Despite accuracy improvements over traditional methods, challenges of interpretability and calibration persist. Therefore, ML outputs serve as decision-support tools rather than independent tools for policy actions.

Natural language processing (NLP) further enriches supervisory insights by extracting real-time sentiment and risk signals from text data. Its role includes quantifying sentiments, creating clarity, and analyzing central bank communications, such as speeches and statements. This enables central banks to measure market reception, identify policy shocks, and refine messaging strategies. However, such transparency comes with risks. NLP tools could allow external actors to interpret policy signals in real-time, adding a new dynamic to central bank communication.

Comparatively, while ML models tend to exhibit superior predictive power over traditional econometric models under certain conditions, they often lack interpretability. To balance these aspects, central banks increasingly adopt hybrid approaches that blend data-driven insights with economic models based on theory.

Central banks worldwide, including the National Bank of Georgia, the European Central Bank, Federal Reserve, and Bank of England are progressively integrating ML and AI into research and policy operations. These applications include inflation and GDP forecasting, financial stability assessments, policy communication analyses, and others. International organizations, including International Monetary Fund and Bank for International Settlements support knowledge sharing initiatives that support emerging market central banks in adopting AI tools for economic nowcasting and risk monitoring, contributing to global cooperation.

However, there are certain challenges in adopting ML/AI for monetary policy. Data quality and availability remain foundational constraints, especially in emerging markets. The interpretability and transparency of complex models are essential for maintaining policy accountability and credibility. ML implementation requires robust infrastructure, governance frameworks, skilled personnel, and protection against cybersecurity threats.

In conclusion, the integration of machine learning and artificial intelligence offers central banks new powerful tools to enhance monetary policy efficiency. These technologies enable faster, richer, and more detailed economic analysis, but require careful, hybrid approach that maintains human judgment and theoretical understanding of decision-making. Responsible adoption, with attention to transparency, robustness, ethical considerations, and global collaboration, will allow monetary authorities to harness AI/ML’s full potential, fostering more resilient, adaptive, and efficient monetary policy frameworks for the 21st century.