AI Regulation Tracker · CFPB · Guidance

What does CFPB Innovation Spotlight on AI/ML adverse action notices (2020) say about AI in banking?

Published Jul 7, 2020 · Last reviewed Aug 26, 2026

In July 2020 the CFPB published an Innovation Spotlight explaining how ECOA and Regulation B accommodate AI and machine-learning underwriting. It noted that Regulation B's Official Interpretations already give creditors flexibility: the sample reasons in Appendix C are illustrative, and a creditor may disclose a reason not on the form if it accurately reflects a factor the model actually used. The post was the Bureau's first written statement that existing adverse-action rules can be met with AI models, and it invited firms to use the Bureau's trial-disclosure and no-action-letter programs to test approaches.

DocumentCFPB Innovation Spotlight on AI/ML adverse action notices (2020)Innovation spotlight: Providing adverse action notices when using AI/ML models
Issued byConsumer Financial Protection Bureau
TypeGuidance
StatusFinal
PublishedJul 7, 2020
Applies toCreditors using or considering AI/ML underwriting models
Official sourceconsumerfinance.gov
Use casesCredit scoring & underwriting · Fair lending & discrimination

What are the key points of CFPB Innovation Spotlight on AI/ML adverse action notices (2020)?

  • Confirms ECOA and Regulation B apply to AI/ML models and that flexibility in the regulation lets creditors disclose reasons not listed on sample forms.
  • Recognizes that AI models may rely on non-traditional data and that reasons must still accurately describe the factors considered.
  • Encourages firms to use the Bureau's then-existing regulatory sandbox, trial disclosure program and no-action-letter policy (later ended in 2022).
  • Now hosted in the CFPB archive; the 2022 circular took a stricter tone on 'black-box' models.
  • Written under Director Kraninger; the Bureau's 2024 comment to Treasury later disavowed sandboxes and no-action letters as unfair advantages.

What did CFPB Innovation Spotlight on AI/ML adverse action notices (2020) change for banks?

It signalled to lenders that AI underwriting was legally viable under existing rules provided the explanation requirement could be met, and set up the framing that later circulars hardened: flexibility on wording, none on accuracy.

Can a lender disclose an adverse-action reason that is not on the Regulation B sample form?

Yes. The sample forms are illustrative; the CFPB's 2020 spotlight and Regulation B's Official Interpretations both say a creditor may and should disclose the actual factor, even if it is not listed.

DateDocumentStatus
Apr 22, 2026Regulation B final rule on disparate impact (April 2026)Equal Credit Opportunity Act (Regulation B) — final rule amending disparate impact, discouragement and special purpose credit program provisionsIn force
May 12, 2025CFPB withdrawal of 67 guidance documents (May 2025)Interpretive Rules, Policy Statements, and Advisory Opinions; WithdrawalIn force
Aug 12, 2024CFPB comment to Treasury on AI in financial services (2024)CFPB Comment on Request for Information on Uses, Opportunities, and Risks of Artificial Intelligence in the Financial Services SectorFinal
Sep 19, 2023CFPB Circular 2023-03Adverse action notification requirements and the proper use of the CFPB's sample forms provided in Regulation BWithdrawn
Jun 6, 2023CFPB Chatbots in Consumer Finance (issue spotlight, 2023)Chatbots in consumer financeFinal
Apr 25, 2023Joint Statement on Automated Systems (CFPB, DOJ, EEOC, FTC)Joint Statement on Enforcement Efforts Against Discrimination and Bias in Automated SystemsFinal

Follow every document these regulators publish

6 curated AI stories for banking executives · Every morning · Free

Subscribe to BankingNewsAI →