Equal Credit Opportunity Act and Regulation B (ECOA / Reg B)

Equal Credit Opportunity Act and Regulation B

In forceECOA / Reg B

Summary

Lenders that deny credit or worsen its terms must give the applicant the specific principal reasons, including when the decision comes from a complex AI model. In May 2025 the CFPB withdrew its circulars on algorithms, but the legal duty is unchanged. A CFPB rule published on April 22, 2026 and effective July 21, 2026 provides that ECOA does not authorize disparate-impact liability. According to secondary sources, it is being challenged in court and does not affect state laws or the Fair Housing Act.

Who it applies to

Creditors extending credit to consumers or businesses in the US, including fintechs and installment-payment platforms.

Penalties

Actual and punitive damages (up to $10,000 in individual actions), class actions and supervisory action.

Official name: Equal Credit Opportunity Act (15 U.S.C. § 1691 et seq.); Regulation B (12 CFR Part 1002)
Official reference: 15 U.S.C. § 1691(d); 12 CFR 1002.9
Regulator: Consumer Financial Protection Bureau (CFPB) and banking regulators
Official source: Check the sourceHow we verify sources

Key dates

May 12, 2025CFPB withdraws Circulars 2022-03 and 2023-03
April 22, 2026CFPB publishes disparate-impact rule (91 FR 21620)
July 21, 2026CFPB rule removing disparate impact under ECOA takes effect

Obligations and dates

Obligations table
Obligation	Article or citation	Status	Date	Who it applies to	Verification
Give specific reasons when AI denies or worsens credit	15 U.S.C. § 1691(d)(2)-(3); 12 CFR 1002.9(a)(2), (b)(2)	In force	Already applicable	Applies if you use AI or scoring models to make credit decisions.	Verified in the official source on November 4, 2026
Test credit models for discrimination	15 U.S.C. § 1691(a); 12 CFR 1002.4	Regulator guidance	Already applicable	Recommended if you use AI in lending, even though the 2026 federal rule narrows disparate-impact liability.	Pending confirmation in the official source (last checked: November 4, 2026)

Give specific reasons when AI denies or worsens credit

15 U.S.C. § 1691(d)(2)-(3); 12 CFR 1002.9(a)(2), (b)(2)In force

Already applicable

When it relates: Applies if you use AI or scoring models to make credit decisions.
What the rule provides: Notify the adverse action within 30 days of a completed application.
State the specific principal reasons for the decision; "failed to meet our internal standards" is not enough.
Be able to translate model variables into accurate, understandable reasons.
Typical evidence: Adverse action notice template
Documented method for deriving reasons from the model
Samples of notices sent
Related base documents: Adverse Automated Decision Explanation Letter
Supporting documents: Model Explainability (XAI) Procedure
Automated Decision Log Standard
Model Risk Management Policy and Model Inventory
Verification: Verified in the official source on November 4, 2026Checked on govinfo (15 U.S.C. § 1691(d)(2)) and eCFR (12 CFR 1002.9(a)(1), 30-day period; 1002.9(b)(2), specific principal reasons). The withdrawal of Circulars 2022-03 and 2023-03 is in the Federal Register (2025-08286, May 12, 2025).

Test credit models for discrimination

15 U.S.C. § 1691(a); 12 CFR 1002.4Regulator guidance