Published document record

EBA Discussion Paper on machine learning for IRB models

PDF source copies

These PDFs are hosted here for continued access. Each is labelled as an unmodified official PDF or a converted reading copy. They are saved copies, not a live statement of the latest law or institutional endorsement. Official source links remain below.

Official PDF — unmodified source copy
If the PDF does not display in your browser, Open PDF in a new tab or download it below. Both use the copy hosted on this site.
Download PDF — 29 pages (1.28 MB)
EBA/DP/2021/04. Discussion paper for consultation.
Source: European Banking Authority. Retrieved 2026-09-10.
Copy provenance and reuse
Unmodified source PDF. Source attribution and original notices are retained; see the source reuse terms. Source reuse terms. The database metadata licence does not replace these terms.
SHA-256: d9e73392782e391ad7dd88cb96f4781cbbab748f486dadf4f95dad5db66444e0

Official metadata

Official title: EBA Discussion Paper on machine learning for IRB models
Official reference: EBA/DP/2021/04
Document type: Consultation document
Document date: 2021-11-11
Document date kind: Document issue
Document-date evidence: European Banking Authority · PDF cover: EBA/DP/2021/04, 11 November 2021. — Exact issue date printed on this version.
Publication date: 2021-11-11
Publication-date evidence: European Banking Authority · Follow-up report p.4 paragraph 3: EBA published and consulted on the discussion paper on 11 November 2021. — Establishes official publication of this version on 2021-11-11, before the 4 September 2026 cutoff.
Version label: Discussion paper for consultation
Legal status: Non binding
Legal-status evidence: European Banking Authority · PDF p.4 Disclaimer: preliminary views do not bind EBA. — Supports classification as non-binding rather than binding legislation.
Language: en

Research classifications

These are researcher classifications used to organise and analyse the corpus; they are not official EU metadata.

Record level

Principal

Version status

Draft

Temporal collection

Contemporary EU AI policy

Relevance class

Direct AI relevance

Policy stage

Consultation

Corpus tier

Directly related

Sectors

Policy membership

Classification evidence

Relevance class: Direct AI relevance · European Banking Authority · PDF pp.4-7: consultation disclaimer, Executive summary and Introduction. — AI/ML is a substantive regulatory or sectoral subject.
Sector tags: Financial services · European Banking Authority · PDF pp.4-7: consultation disclaimer, Executive summary and Introduction. — Addresses regulated banking, insurance or securities activities.
Provenance tags: Authored by an EU agency or body · European Banking Authority · PDF pp.4-7: consultation disclaimer, Executive summary and Introduction. — Produced within the named EU agency/body; named personal author credits remain separate.
Provenance tags: Officially published · European Banking Authority · Official record with linked English document. — Official EU endpoint distributes this independently citable English text.
Provenance tags: Official consultation material · European Banking Authority · Official consultation status and linked draft/discussion paper. — Formally published consultation material.

Production provenance

Provenance tags

Named bibliographic authors

European Banking Authority · European Banking Authority · PDF institutional title and imprint.

Institutions and roles

European Banking Authority · Author · role evidence · PDF institutional title/imprint and substantive text.
European Banking Authority · Official host · role evidence · Official publication record and linked English PDF.
European Banking Authority · Publisher · role evidence · Official publication record and linked English PDF.

Official sources and identifiers

Official source · Publisher/host: European Banking Authority
Official source · Publisher/host: European Banking Authority
Official source · Publisher/host: European Banking Authority

Research assessment

Inclusion rationale: Direct banking-supervision AI/ML consultation document and the explicit precursor to the known 2023 follow-up report.
Researcher notes: Consultation document with expressly preliminary, non-binding views. The follow-up report independently confirms publication on 11 November 2021.
Review status: verified

Relationships

EBA Discussion Paper on machine learning for IRB models · Based on · EBA Report on Big Data and Advanced Analytics

Official relationship · Verified

Official rationale: Discussion paper p.7 Introduction explicitly builds on the 13 January 2020 BD&AA report and its ML framing.

Evidence source: European Banking Authority
Machine Learning for IRB Models: Follow-up report from the consultation on the Discussion Paper on machine learning for IRB models · Based on · EBA Discussion Paper on machine learning for IRB models

Official relationship · Verified

Official rationale: Follow-up report pp.3-5 summarises consultation on the 11 November 2021 discussion paper and retains/refines its recommendations.

Evidence source: European Banking Authority

Verification

Expanded evidence review: Verified

Reviewed by: Yichen Hao
Evidence review date: 2026-09-05T23:48:11Z

The evidence review date is not a separate human sign-off date. Review credit and evidence-review actions

Last verified: 2026-09-05T23:48:11Z · PDF cover: EBA/DP/2021/04, 11 November 2021. Official linked English manifestation reviewed.
Last verified: 2026-09-05T23:48:11Z · PDF pp.4-7: consultation disclaimer, Executive summary and Introduction. English text inspected in web extraction or Node fetch plus pypdf; no retained file snapshot claimed.
Last verified: 2026-09-05T23:48:11Z · PDF pp.3-5, summary and consultation history; section 5.2 and Annex I. English text inspected in web extraction or Node fetch plus pypdf; no retained file snapshot claimed.

Return to the Corpus