ECOA’s Safe Harbor. Or, What If the CFPB Is Wrong on Regulation B?
Aug 20 2026 • 5 Min Read
The Consumer Financial Protection Bureau 's 2026 amendments to Regulation B removed the regulation's longstanding effects-test language and adopted the position that the Equal Credit Opportunity Act (ECOA) does not impose liability for disparate-impact on a prohibited basis. Shortly before the rule's effective date, several advocacy organizations filed suit challenging the amendments under the Administrative Procedure Act and ECOA, seeking vacatur of the rule.
However, the revisions to Regulation B were not stayed and became effective on July 21, 2026. As the litigation proceeds, much of the discussion has centered on whether the CFPB's rule will ultimately survive judicial review, focused largely on whether the ECOA permits disparate-impact claims.
But the validity of the CFPB's interpretation is not the only issue that matters for creditors. Post Loper Bright, a court may conclude that ECOA authorizes disparate-impact claims notwithstanding the CFPB's contrary interpretation. In that circumstance, however, ECOA contains a statutory safe harbor that may substantially limit liability arising from conduct undertaken while the revised rule remained in effect. Section 1691e(e) protects creditors for any act “done or omitted” in good-faith conformity with official CFPB rules, regulations, or interpretations, even when those authorities are later amended, rescinded, or determined by judicial authority to be invalid for any reason.
A rule that is later vacated is not the same thing as a rule that never took effect. Because the court did not stay the amendment’s effective date, the revisions replaced the previous rule on July 21, 2026. Unless a court orders otherwise or a new administration makes further revisions, the Bureau’s “official regulation” does not contemplate disparate impact. However, because a court may hold otherwise or a new administration may revise Regulation B again, creditors should consider the scope of ECOA's safe-harbor provision.
Private plaintiffs, federal agencies, or state officials may assert disparate-impact claims arising from conduct that occurred while the amended rule was in effect, arguing that a reviewing court should reject the CFPB's interpretation and recognize disparate-impact liability under ECOA. The courts in those cases may agree. Additionally, the plaintiffs in the current suit challenging the Regulation B rulemaking may win. The court in that case may conclude that the CFPB's interpretation was incorrect, that ECOA authorizes disparate-impact claims, and that the 2026 amendments must be vacated. Those conclusions would, to an extent, resolve the statutory question, but not the liability question.
ECOA's safe harbor is a statutory provision. It seems unlikely that a court would use its independent judgment to reject an agency’s reading of the statute while ignoring an express statutory defense enacted by Congress to protect parties that rely on the Bureau’s official rules. As a result, a court that concludes ECOA authorizes disparate-impact liability should determine that § 1691e(e) bars liability for conduct undertaken in good-faith conformity with the amended Regulation B. If so, the revival of disparate-impact liability would be prospective, governing future conduct only. Conduct occurring during the rule's effective period would be evaluated under § 1691e(e). By the time appellate review is complete, limitations defenses may further restrict available claims. As a result, the combined effect of the statutory safe harbor and ordinary limitations principles could substantially narrow the universe of actionable disparate-impact claims, leaving any future disparate-impact regime entirely focused on future conduct.
The principal uncertainty surrounding § 1691e(e) is what it means to act "in good faith" and "in conformity" with the CFPB's amended regulation. Most safe-harbor decisions arise in a very different context. The typical case involves an agency rule, interpretation, model form, or official commentary that affirmatively prescribes or authorizes particular conduct. Courts then compare the defendant's conduct against the requirements of that guidance and ask whether the defendant actually followed it.
ECOA cases applying the safe harbor generally fit that pattern. Courts have found protection where a creditor used adverse-action language that tracked or substantially mirrored model forms and official guidance issued under Regulation B. In those circumstances, the conformity inquiry is straightforward because the court can compare the creditor's conduct to specific language supplied by the agency. Conversely, courts considering analogous safe harbors under other federal consumer-finance statutes have emphasized that protection is unavailable where a regulated party departs from the governing guidance, stretches it beyond its terms, relies on guidance that does not actually address the conduct at issue, or merely forms its own mistaken view of what the law requires.
Federal courts applying safe-harbor provisions have reached much the same conclusion. Courts generally do not treat subjective belief alone as sufficient. Instead, they focus on whether the defendant can demonstrate actual reliance on qualifying agency guidance and actual conformity with that guidance. Safe harbors are commonly denied where regulated entities ignore conditions embedded in agency interpretations, apply those interpretations outside the circumstances to which they were directed, or rely on guidance they did not in fact follow. The recurring question is whether the challenged conduct can be paired with a specific regulatory requirement, authorization, or instruction, and whether the defendant acted consistently with that requirement in objective good faith.
The CFPB's 2026 revisions present a different situation. The amended rule does not establish procedures, prescribe disclosures, require monitoring, mandate underwriting practices, or otherwise direct creditors to take affirmative action with respect to disparate impact. It imposes no particular directive. Instead, the rule removes Regulation B's effects-test language and states that ECOA does not recognize disparate-impact liability. As a result, the usual conformity analysis seems out of place. There is no model form to follow, no required disclosure to provide, and no prescribed compliance step against which a creditor's conduct can be measured. A creditor invoking § 1691e(e) likely would argue that “conformity” consists simply of operating under a regulatory regime in which disparate-impact liability is not a governing legal standard.
Future potential plaintiffs likely would want to advance a narrower view. They may argue that § 1691e(e) requires something more than a favorable regulation sitting on the books. But because the CFPB's rule does not impose conditions that can be followed or violated, ECOA’s safe harbor in this context may be broader than in prior, more traditional cases.
The practical takeaway is that ECOA provides a potentially substantial statutory defense to disparate-impact claims arising from conduct undertaken while the rule remains in effect. That defense, however, is limited to disparate-impact liability. The revisions to Regulation B did not affect ECOA's prohibition on intentional discrimination, disparate-treatment and proxy-discrimination theories. Additionally, the Fair Housing Act and state fair lending requirements remain in place. Accordingly, institutions should maintain risk-management, testing, monitoring, and governance practices appropriate to their fair-lending obligations.