The EBA publishes follow-up Report on the use of machine learning for internal ratings-based models

The European Banking Authority (EBA) published today a follow-up Report presenting the feedback received during the consultation on machine learning (ML) used in the context of internal ratings-based (IRB) models. This follow-up Report summarises the main conclusions from the consultation and provides an overview of the current use cases of ML techniques for IRB models.

In November 2021, the EBA published a discussion paper, seeking feedback on the use of ML in the context of IRB models, focussing on the challenges and opportunities faced by practitioners.

In this Report, the EBA reiterated its recommendations proposed in the discussion paper, which were broadly supported by the respondents, aiming at ensuring a prudent use of machine learning models in the context of the IRB framework.

The use of ML techniques in credit risk models may create issues beyond the scope of prudential consideration. In this respect, this follow-up Report also discussed the interaction with two other legal frameworks, namely the General Data Protection Regulation (GDPR) and the Artificial Intelligence (AI) Act and called for some clarifications in order to reduce legal uncertainty and avoid unintended consequences of the AI Act.