In the field of supervision, artificial intelligence has also proven to be useful in monitoring financial stability, an essential part of the work of central banks globally. One of the most promising approaches is the use of large language models (LLM) to quantitatively assess the effectiveness of communication from monetary authorities. Sharing experiences on successful implementations and promoting collaborative research will be key to maximizing the benefits of these technologies.
José Luis Escrivá, a representative of the Bank of Spain, presented some of the challenges facing central banks in the digital era. Among the most notable current applications is the analysis of macroeconomic data, such as the use of high-frequency information to monitor price signals and develop leading indicators on economic activity and financial flows. Establishing an interaction space for economists interested in issues related to monetary policy in Latin America is the purpose of this initiative.
To fully capitalize on the potential of artificial intelligence, it is vital to strengthen human capital in the research teams of central banks by incorporating AI specialists focused on monitoring and economic evaluation. Additionally, ensuring adequate technological infrastructure, such as cloud computing and specialized microprocessors, is essential for optimizing data processing. The quality and structuring of the data also play a crucial role in the precision of models and in reducing their learning time.
Amid the global trend toward fragmentation, it is crucial to maintain and strengthen international cooperation among central banks. Machine learning models are complementing traditional prediction approaches by enabling real-time forecasts based on the analysis of large amounts of information. Another significant advancement is the use of high-frequency data monitors to assess the impact of events on financial markets.
In the specific field of central banks, artificial intelligence is being used to enhance models based on numerical data by integrating text analysis of statements and minutes from monetary authorities. The intention is to generate a dialogue space similar to the well-known annual Jackson Hole event, where the monetary policy of the Federal Reserve of the United States is discussed, but in the Latin American context.
The Chapultepec conference, organized by the Americas office of the Bank for International Settlements (BIS), based regionally in Mexico, provided a platform to explore increasingly relevant topics in the analysis of regional monetary policy, such as digitization, technology, cryptocurrencies, and the early applications of artificial intelligence in specific tasks of central banks. Although the use of AI in central banks is still in development, its potential to improve accountability mechanisms is recognized, although it also faces challenges in its implementation.
