In June, the Bank for International Settlements (BIS)—known as the “Bank of Central Banks”—urged countries to use Artificial Intelligence (AI) to develop analytical tools for their central banks. These tools would, among other things, track real-time data to help predict inflationary trends.
Such monitoring of economic indicators would have been useful in the aftermath of the pandemic or of Russia’s invasion of Ukraine when big central banks failed to grasp just how far and how fast inflation would subsequently rise.
The BIS Annual Report 2024 highlights the impact of new AI applications for central banks, pointing out that the technology is expected to affect the financial system, reshape markets, influence production trends, and impact economic growth. AI is helping firms adjust their prices more rapidly in response to macroeconomic changes, for instance, with implications for inflation dynamics.
AI’s benefits and risks
The financial sector could be one of the biggest beneficiaries of AI if it enhances its efficiency in lending, financial intermediation, insurance, forecasting, asset management, and monitoring liquidity and payments.
AI promises to reduce the cost of time-consuming tasks traditionally done by people, improve customer service, improve fraud detection, and ensure regulatory compliance, such as with the Know Your Customer (KYC) protocols, and with anti-money laundering and counter-terrorist financing systems.
The picture is not all rosy, however. AI could increase risks in the financial sector, such as with biased data or flawed models. If AI tools are trained on out-of-date data, this could result in distorted forecasts and decisions There are also worries about systems' exposure to complex cybercrime.
Gary Gensler, chair of the US Securities and Exchange Commission (SEC) and a professor at the Massachusetts Institute of Technology (MIT), speaks of another issue, warning of the dangers of 'herd behaviour', with algorithms unwittingly provoking or exacerbating financial crises.
Finally, while AI will render some financial jobs redundant, new technology always creates new jobs. Yet the risk with this technology is of concentrating its understanding and operation in the hands of a small number of specialists.
Using AI to manage data
Central banks are the crossroads of the monetary, financial, and banking systems. Custodians of the economy through their monetary policy mandate, they play a key role in maintaining financial stability and regulating and supervising commercial banks and other financial institutions.
Due to the high proportion of central bank tasks that require data-intensive cognitive skills, they are in a privileged position to reap the benefits of AI and Machine Learning tools to support their core functions, such as setting monetary policy, managing the payments system, and collecting and analysing data—their primary resource.
The importance of data has grown with the advent of AI and its growing ability to ingest huge amounts, churning it into information. The question facing central banks is no longer whether to use AI but rather how to use it effectively and responsibly.
Ideally, AI will help to make data (whether structured or unstructured) available in a timely manner and organise it in a way that produces accurate assessments and real-time forecasts that can be easily updated as new data comes in, whether this be on growth, inflation, employment, production, or supply and demand.
Since a single indicator is usually insufficient to accurately track economic activity in real-time, forecasting models often process diverse and complex datasets to develop early warning indicators for potential stress points that could pose systemic risks.
Technology and governance
Beyond the challenge of rapid, real-time data collection, central banks also need to master data governance, specifically the reformatting of data into useful models. Using external models may be more cost-effective, but the reliance on a small number of providers exposes central banks to risk.
Today, data is mainly in the hands of companies who offer related services that are widely used by central banks. The cost of this has increased significantly in recent years, and providers have imposed stricter terms of use.
