# EBA follow-up report on machine learning for IRB models (EBA/REP/2023/28): Machine Learning for IRB Models — Follow-up report from the consultation on the Discussion paper on machine learning for IRB models

Source: https://www.bankingnewsai.com/ai-regulation/documents/eba-rep-2023-28-ml-irb-follow-up
Last updated: Aug 26, 2026

The EBA's follow-up report on machine learning for IRB models (EBA/REP/2023/28), published August 4, 2023, closed out the 2021 consultation. It confirmed that respondents broadly supported the EBA's principle-based recommendations for prudent ML use in IRB models and reiterated them, mapped the ML use cases banks actually report — ML in core modelling steps, ML challenger models in validation, and collateral valuation — and flagged interactions with GDPR and the then-draft AI Act, asking for clarifications to reduce legal uncertainty for capital models.

## At a glance

| Field | Value |
| --- | --- |
| Authority | [EBA](https://www.bankingnewsai.com/ai-regulation/eba) |
| Type | Report |
| Status | Final |
| Published | Aug 4, 2023 |
| Applies to | EU credit institutions using machine learning in IRB credit-risk models; relevant to ECB and national supervisors approving such models |
| Official text | https://www.eba.europa.eu/publications-and-media/press-releases/eba-publishes-follow-report-use-machine-learning-internal |

## Key points

- Published August 4, 2023 as EBA/REP/2023/28.
- Current use is selective: ML in core IRB modelling steps, ML as challenger models for validation, and collateral valuation, rather than fully ML-based rating systems.
- Complexity issues addressed: overfitting, human-skill requirements, and explainability of individual outputs.
- Reiterates the discussion paper's recommendations on understanding, documentation, validation and stability, which respondents broadly supported.
- Discusses interaction with GDPR (automated decision-making) and the AI Act, calling for clarification to avoid unintended consequences for IRB models.
- No new guidelines were issued; the report is the standing statement of EBA expectations for ML in capital models.

## What changed for banks

The follow-up turned a discussion into a settled position: ML in IRB models is acceptable within existing CRR requirements, and banks should expect supervisors to test explainability, understanding and stability rather than prohibit the technique. It also put on record, a year before the AI Act was adopted, that credit-scoring models used for capital purposes could be caught by the Act — an issue the EBA's 2025 mapping exercise later addressed.

## Use cases it governs

- [Model risk management](https://www.bankingnewsai.com/ai-regulation/by-use-case#model-risk)
- [Credit scoring & underwriting](https://www.bankingnewsai.com/ai-regulation/by-use-case#credit-underwriting)

Supersedes: [EBA discussion paper on machine learning for IRB models](https://www.bankingnewsai.com/ai-regulation/documents/eba-ml-irb-discussion-paper-2021).

## FAQ

### What did the EBA conclude on ML in IRB models in 2023?

In EBA/REP/2023/28 (August 4, 2023) the EBA reiterated its 2021 recommendations for prudent use — management understanding, documentation, explainability, control of overfitting and model stability — noted that current use is mostly in core modelling steps, challenger models and collateral valuation, and asked for clarity on GDPR and AI Act interactions.

### Does the EBA ML-for-IRB report change capital rules?

No. It sets expectations on how existing CRR IRB requirements apply when ML is used; it did not amend the CRR or issue new guidelines.

## Related documents

- [EBA discussion paper on machine learning for IRB models](https://www.bankingnewsai.com/ai-regulation/documents/eba-ml-irb-discussion-paper-2021) — Discussion Paper on machine learning for IRB models (Nov 11, 2021)
- [ESA Statement on ICT risks from frontier AI models (JC 2026 25)](https://www.bankingnewsai.com/ai-regulation/documents/esas-jc-2026-25-frontier-ai-statement) — ESA Statement: Toward a consistent and risk-based approach for ICT risks from frontier AI models (Jul 31, 2026)
- [EBA factsheet: AI Act implications for the EU banking and payments sector](https://www.bankingnewsai.com/ai-regulation/documents/eba-ai-act-factsheet-2025) — AI Act: implications for the EU banking and payments sector (Nov 21, 2025)
- [EBA Chair letter to the Commission on the AI Act mapping exercise (EBA/2025/D/5384)](https://www.bankingnewsai.com/ai-regulation/documents/eba-2025-d-5384-ai-act-mapping-letter) — Outcome of EBA's AI Act mapping exercise — letter to DG FISMA and DG CNECT (Nov 21, 2025)
- [EBA Work Programme 2026](https://www.bankingnewsai.com/ai-regulation/documents/eba-work-programme-2026) — EBA Work Programme 2026 — AI Act implementation and digital-finance priorities (Oct 1, 2025)
- [EBA report: Rising application of AI in EU banking and payments (Sep 2025)](https://www.bankingnewsai.com/ai-regulation/documents/eba-ai-adoption-report-2025) — Rising application of AI in EU banking and payments sector (Sep 25, 2025)
- [EBA/GL/2022/15 (remote customer onboarding)](https://www.bankingnewsai.com/ai-regulation/documents/eba-gl-2022-15-remote-customer-onboarding) — EBA/GL/2022/15 Guidelines on the use of Remote Customer Onboarding Solutions under Article 13(1) of Directive (EU) 2015/849 (Nov 22, 2022)
- [EBA Guidelines on loan origination and monitoring (EBA/GL/2020/06)](https://www.bankingnewsai.com/ai-regulation/documents/eba-gl-2020-06-loan-origination-monitoring) — Guidelines on loan origination and monitoring (May 29, 2020)

Last reviewed Aug 26, 2026. Cite the official text (https://www.eba.europa.eu/publications-and-media/press-releases/eba-publishes-follow-report-use-machine-learning-internal) for the rule and this page for the summary and dates.

---

Canonical page: https://www.bankingnewsai.com/ai-regulation/documents/eba-rep-2023-28-ml-irb-follow-up
Part of [BankingNewsAI](https://www.bankingnewsai.com/) — a free daily brief on AI in banking, an AI regulation tracker (41 authorities, 237 documents) and AI-strategy profiles of the 120 largest US banks. Markdown versions of every reference page: append `.md` to the page URL; index at https://www.bankingnewsai.com/llms.txt.