Rising inflation poses a significant global challenge, impacting households, businesses, and governments alike. Central banks play a crucial role in addressing this by stabilizing the economy through the delicate balance of adjusting interest rates and managing the money supply.

In support of this mandate, many central banks worldwide became early adopters of AI and machine learning techniques, leveraging this technology for macro analysis, forecasting, payment systems oversight, statistics, and more. But as AI evolves, it’s time for central banks to hone their approach, unlocking AI’s full potential while applying a global, collaborative lens to its use for policy decisions.

Getting this right is vital, as it promises significant efficiency gains with far-reaching global benefits. However, achieving these outcomes won’t be straightforward; it requires a carefully crafted blueprint for the ongoing use and development of AI.

This blog, the first in a two-part series, will assess what the blueprint should look like.

How Are Advances in AI Transforming the Role of Central Banks?

While the key role of central banks remains the same, the rapidly advancing capabilities of AI can transform how these institutions fulfill these critical objectives by:

• Processing and analyzing economic data at speed and scale to enhance nowcasting and forecasting, financial stability monitoring, and monetary policy implementation. Advanced algorithms can enable real-time monitoring of activities and markets, identifying patterns and trends to predict inflation, unemployment, and other economic indicators more accurately and quickly than traditional models. This also allows central banks to respond more swiftly to changes in market conditions, particularly during periods of heightened market volatility. 
• Improving productivity, operational efficiency, and regulatory compliance. AI can automate routine tasks such as data entry, document processing, and report generation, vastly improving efficiency. This strengthens regulatory compliance by effectively detecting anomalies and potential issues. Additionally, AI capabilities can provide decision-making support for complex tasks, generating actionable insights from data and enabling exception handling.
• Supporting the design, development, and regulation of new technologies and innovations. These run the gamut from payment services and systems to lending, wallets, investment platforms, and digital currencies. AI can assist central banks in understanding, monitoring, and overseeing new platforms and solutions in a more streamlined, secure way.
• Protecting consumers and detecting fraud. Intelligent capabilities augment the detection of fraudulent activities by analyzing transaction data and identifying unusual patterns. This safeguards consumers and the wider financial system. 

These developments are just the start. The body of research and emerging use cases is expanding rapidly as more central banks around the world deepen their technological investments and experiment with the expanding capabilities of AI to become more agile, informed, and efficient in managing economic and financial stability.

AI in Action

Piero Cipollone, an executive board member of the European Central Bank, recently outlined how the ECB is using AI in their operations. 

Six years ago, the ECB first began applying AI to improve the efficiency, accuracy, and quality of its statistical processes. Now, the bank’s use of AI has evolved to include machine learning models and large language models (LLMs) that nowcast and forecast inflation – using a mix of text data and ML techniques to quantify risks and tensions in the global economy. 

AI also scans millions of documents to help regulators spot anomalies. It further translates over six million pages a year in all 24 languages of the EU – far greater than the 150,000 pages the ECB’s language services would be limited to covering without those tools.

Harnessing, streamlining, and optimizing economic forecasting and operations will allow central banks to spot anomalies, overcome challenges, and turn them into opportunities to proactively manage and resolve threats through more informed and timely policy decision-making.

Balancing AI and Human Judgment in Policy Decisions

Combining AI’s efficiency with human intelligence is the key to unlocking greater value, making it the next critical stage of the roadmap. AI capabilities alone cannot replace the value and impact of human judgment, particularly when it comes to making critical and nuanced monetary policy decisions.

The economy is influenced by various factors like social, political, and global events that are too complex to quantify and model. Human policymakers can consider these qualitative aspects and understand the broader context in which economic data exists.

Policy decisions often involve choices about what’s important and right, which aren’t easily measurable. Take, for example, deciding between controlling inflation and keeping unemployment low. This requires understanding what society values more against the backdrop of current socio-economic events. 

Sometimes, economic outcomes also have consequences for international relationships and trade, which is difficult for an AI model to consider. What this means is that while AI can be used to model various scenarios, human judgment is needed to interpret them. 

Augmenting Capabilities with a Human-In-The-Loop Approach

Interest rate decision-making, for instance, often involves consensus-building among different stakeholders, including government officials, financial market participants, and the public. Policymakers are better equipped to engage with these groups and address their concerns with the necessary transparency, empathy, and accountability.

Not all scenarios are clear-cut or have a single optimal solution, either. Human decision-makers can navigate ambiguous situations by weighing different considerations and making informed choices even in the absence of complete data. 

Human judgment also allows for flexibility and adaptability in policy responses, as policymakers can adjust their strategies based on new information and changing circumstances, as well as key ethical and regulatory considerations.

AI augments human capabilities in making these decisions. AI can handle large-scale data processing and analysis efficiently, while human experts contribute their experience, intuition, and understanding of complex economic conditions. This combination ensures that decisions are data-informed, contextually relevant, and address any hallucination effects of technology.

To ensure that human oversight complements AI-driven insights, central banks will need to establish clear governance frameworks that define the roles and responsibilities of both human experts and AI systems. This includes having dedicated teams for overseeing AI applications and ensuring they align with policy objectives. They will additionally need to invest in training staff on how to interpret and use AI-generated insights. 

It will also become essential for central banks to conduct regular audits and create feedback loops to refine AI models to become exponentially more accurate and relevant over time. By ensuring that AI systems are transparent and their outputs are interpretable, human policymakers can better evaluate and trust the AI’s recommendations.

Increasingly, that means harnessing a ‘Human-in-the-Loop’ training model – which combines advanced technological capabilities with human expertise and improves the learning of humans and AI at the same time for cumulatively better outcomes. Journalist and Forbes contributor Mike O’Sullivan calls it the ‘One Man and his Dog’ approach, where humans perform a complex task under pressure with the aid of a trained, intelligent non-human – previously a dog, and now AI. 

The more we can streamline the combined strengths of humans and advanced platforms, the greater the benefits we can deliver through a balanced approach that enhances decision-making and policy effectiveness.

Looking Ahead: The Future of AI in Central Banking and Finance

As AI continues to advance, its impact on the future of central banking is undeniable. It will revolutionize the way central banks analyze data, implement monetary policies, and maintain financial stability. 

The integration of AI into their operations will not only drive unprecedented efficiency but also enhance predictive accuracy, strengthen risk management, and refine decision-making processes. Furthermore, AI will bolster regulatory oversight and improve the ability to respond swiftly to economic fluctuations and systemic threats.

However, the true game-changers will be those forward-thinking central banks that proactively harness AI’s potential, embracing opportunities and mitigating risks. These institutions will lead the way in delivering transformative outcomes, setting new benchmarks for the future of global finance.