Skip to content
Compliance & Risk

European Central Bank sees supervisory boost from generative AI

Trond Vagen  European Correspondent / Thomson Reuters Regulatory Intelligence

· 5 minute read

Trond Vagen  European Correspondent / Thomson Reuters Regulatory Intelligence

· 5 minute read

In the past three years, the ECB has developed 14 applications and platforms serving more than 3,500 users across the ECB and EU members’ banking supervisory authorities

The European Central Bank (ECB) has found more than 40 use cases for generative artificial intelligence (Gen AI) in banking supervision, highlighting the potential for new technology to make day-to-day supervision easier, a senior official said.

Among the more promising use cases was a tool that could translate queries written in plain language into code, helping users find specific data points that previously would have required coding knowledge, said Elizabeth McCaul, a member of the ECB supervisory board, in a blog post published on the ECB website.

McCaul said the ECB operates a “data lake” for European banking supervision that requires some understanding of programming to access. “But with the help of generative AI, which can automatically translate natural language queries into scripts, supervisors with no programming experience can ask where to find very specific data points,” she said. “This is just one example of how generative AI, and AI in general, can make traditional technologies easier to use and put supervisory technology at supervisors’ fingertips.”

The ECB has also developed several proofs-of-concept demonstrating the potential of Gen AI, she added, including one that instantaneously retrieves answers to questions on supervisory methodology, with clear references to internal methodologies.

Gen AI has already begun sweeping through Europe’s financial services sector, according to an October 2023 survey report by Big Four consultancy EY. In fact, 60% of survey respondents said their firm had invested in Gen AI technologies over the past 6 to 12 months, and 75% of executives planned to accelerate spending on these technologies over the next year, the EY European Financial Services AI Survey found. Further, 77% of respondents said they were bracing for a significant AI impact on their workforce or operations.


Gen AI has already begun sweeping through Europe’s financial services sector, according to an October 2023 survey report by Big Four consultancy EY.


As one of the world’s most important central banks, the ECB stands to influence Gen AI’s growth in the sector even further. Its AI focus resulted from previous investments in supervisory technology and a cloud-based virtual lab that features machine learning capabilities and an environment for sharing and developing code, McCaul explained.

“At an early stage, we recognized the need to embrace digital innovation and AI to make European banking supervision more efficient and effective,” she said. “We introduced an ambitious digital agenda to enhance our analytical capabilities. We invested in a portfolio of supervisory technology applications to supervise a complex banking sector and manage an ever-expanding pool of data and tasks.”

Over the past three years, the ECB has developed 14 applications and platforms serving more than 3,500 users across the ECB and the European Union members’ banking supervisory authorities, she said. “Our AI applications enable us to query supervisory data and employ chatbot functionalities for supervisory regulations and methodologies.”

Embracing digital innovation

For textual analysis, the ECB’s Athena application translates and analyzes the content of supervisory documents and combines it with information from other sources, such as public media.

In big data analytics, a tool called GABI can generate and optimize regression models on a large scale, enabling supervisors to base their assessments on a much larger set of models, McCaul said. Further, in the field of network analysis, a tool called NAVI can generate network diagrams to illustrate relationships in data. “This enables users to visualize the often-complex ownership structures of supervised banks and combines data from numerous sources to provide a comprehensive overview of bank owners and interdependencies,” she added.

Other supervisory technologies include Heimdall, which supports the processing of vast amounts of information received to assess senior managers’ fitness and propriety, and Medusa, which facilitates the drafting and consistency checks of reports following internal model investigations.

The ECB has invested in multiple advanced capabilities, hoping to leverage the voluminous data it collects to spot risks at an earlier stage. These investments mirror the broader prioritization of AI development across the European Union. In December, for instance, EU lawmakers agreed on a provisional deal for the EU AI Act, which aims to govern the use of AI within EU member countries and to regulate the region’s major AI-enabled systems, such as ChatGPT.

Thierry Breton, the European Commissioner for the Internal Market of the EU, said in a statement that the bloc’s AI Act will take a “two-tiered approach” to regulation, “with transparency requirements for all general-purpose AI models and stronger requirements for powerful models with systemic impacts across our EU Single Market.”


Zach Warren of the Thomson Reuters Institute contributed to this blog post.