Authors : Jérôme Coffinet, Jean-Brieux Delbos

Bulletin No. 266, article 6. Artificial intelligence (AI) has become a key issue for central banks, both because of its impact on the economy and because of the opportunities it offers in the performance of their missions. 

Indeed, central banks did not wait for the recent boom in generative AI – notably ChatGPT, Gemini, Mistral, Claude and DeepSeek – to begin experimenting with solutions based on big data, data science and AI. They monitor technological developments and embrace AI to leverage it to the full across their various business areas and manage the associated risks.

Drawing on examples from the Banque de France and the Autorité de contrôle prudentiel et de résolution (ACPR – Prudential Supervision and Resolution Authority), this article shows how AI has become a strategic driver of transformation for central banks and financial supervision, both in Europe and the rest of the world.

Number of articles containing the keywords "artificial intelligence" and "central banks" or "central banking"
Number of articles containing the keywords "artificial intelligence" and "central banks" or "central banking" Period 2000-25

1 Central banks are embracing artificial intelligence 

Combined with the massive volumes of data required to train models, artificial intelligence (AI) constitutes a disruptive innovation with considerable potential. It has become a powerful driver of innovation, both for central banks’ internal processes and for transforming their ecosystem. Alongside analytical models designed for specific tasks, there are now generative AI algorithms capable of mimicking human reasoning. Thanks to large language models (LLMs), a query phrased in natural language rapidly elicits an automated response. Their development since 2022 has facilitated human machine interaction. 

The rise of AI has fascinated the public, with the emergence of generative AI tools such as ChatGPT in late 2022, followed notably by Gemini, Mistral, Claude and DeepSeek in 2023. A global survey (Singla et al., 2025) reports rapid take-up: in July 2024, 71% of companies were using generative AI in at least one business function, compared with 33% in 2023. An internal survey conducted by the Banque de France in 2025 concluded that 45% of respondents were using AI assistants less than a year after their launch. 

Central banks had already started to explore the fields of big data, data science and AI long before the rise of generative AI. They drew on machine learning and deep learning (deep neural networks), classification and natural language processing (NLP). Between 2000 and 2025, the increase in the number of articles indexed by Google Scholar reflects the growth in academic research into AI related to central banks (see chart).

Generative AI has vast potential: Korinek (2023) identifies more than 25 use cases for economists (ideation, modelling, agent-based simulation, generation of fictitious data, etc.), most of which can be applied to other professions. 

According to surveys by the Bank for International Settlements (BIS; Araujo et al., 2025), central banks are currently focusing their use of AI on high value-added areas: chatbots, document drafting and summary, code generation, economic research, and statistical production. In the medium term, the greatest benefits are expected in cybersecurity (enhanced threat detection and risk monitoring), as well as in data analysis, forecasting and process automation (see Box 1). AI offers a means of boosting productivity and automating repetitive tasks. More broadly, it is seen as a tool for transforming and accelerating data use, as well as enhancing decision making and operational efficiency.

Updated on the 18th of November 2026