Tag: ECOA

  • The Borrowed Notice: How a Six-Week Repeal, a Vanished Bank Exemption, and a Rulemaking Hearing on Its Final Comment Day Made Colorado the First State AI Law to Bind Banks Directly

    The Borrowed Notice: How a Six-Week Repeal, a Vanished Bank Exemption, and a Rulemaking Hearing on Its Final Comment Day Made Colorado the First State AI Law to Bind Banks Directly

    Colorado repealed its first-in-the-nation AI law on May 14, 2026 — six weeks before it would have taken effect — and replaced it with a disclosure regime that quietly drops the exemption that kept banks and credit unions outside its scope. The centerpiece trade: a lender’s existing federal adverse-action notice now does double duty as state compliance, while the anti-bias obligations that once sat on both sides of that trade are gone.


    A Six-Week Repeal

    In 2024, Colorado enacted Senate Bill 24-205, the first comprehensive state AI law in the United States, built around “high-risk artificial intelligence systems” and a duty to prevent algorithmic discrimination.[1] It was set to take effect June 30, 2026. On May 14, 2026 — with days left in the legislative session — Governor Jared Polis signed Senate Bill 26-189, repealing and replacing the statute “mere weeks before the original statute’s June 30, 2026 effective date,” in the words of one legal analysis.[2] The replacement law, the Automated Decision-Making Technology (ADMT) Act, takes effect January 1, 2027.[2]

    The rewrite changes the law’s basic vocabulary. Gone are “high-risk AI systems”; in their place is “automated decision-making technology” — any technology that processes personal data and uses computation to generate outputs used to make, guide, or assist a decision about an individual.[1] The law applies only when ADMT “materially influences” a “consequential decision,” meaning the system’s output is a non-de minimis factor that affects the outcome, including by ranking, scoring, or recommending.[2] Consequential decisions include access to or eligibility for education, employment, housing, healthcare, insurance, essential government services — and financial or lending services.[1]

    The design logic behind the swap is a bet on regulation by disclosure rather than regulation by governance. Instead of requiring firms to test systems for bias before deployment and maintain formal risk programs, the state requires them to tell consumers when ADMT shaped a decision and give them a route to challenge it after the fact.[2] Whether post-decision transparency can substitute for upstream testing — or merely relocates the burden from institutions to consumers — is the question the rest of the law turns on.


    What Died, What Survived

    The replacement eliminates most of the 2024 law’s substantive obligations:[2]

    • The anti-discrimination duty. Developers and deployers no longer must use “reasonable care” to protect consumers from algorithmic discrimination — the concept the 2024 law was built around. The replacement statute “does not mention algorithmic discrimination.”[2]
    • Impact assessments and risk management programs. Deployers no longer complete annual impact assessments or maintain formal risk-management policies for covered systems.[2]
    • Attorney general reporting. The 90-day duty to notify the attorney general after discovering algorithmic discrimination is gone.[2]
    • The bank and credit union exemption. The provision letting regulated financial institutions claim full compliance through prudential oversight is eliminated entirely.[3]

    What survives is a procedural shell: deployers must give clear, conspicuous notice at the point of interaction; provide a plain-language description of the decision and the ADMT’s role within 30 days of an adverse outcome; honor consumer rights to access and correct inaccurate personal data; offer meaningful human review and reconsideration “to the extent commercially reasonable”; and retain compliance records for three years.[1][2] Enforcement belongs exclusively to the attorney general under the Colorado Consumer Protection Act, where a violation is a deceptive trade practice, with a 60-day cure opportunity before January 1, 2030 when a cure is possible — and no new private right of action.[1]


    The Exemption That Vanished

    For banks and credit unions, the change cuts in a direction most coverage of the repeal missed. The 2024 law contained an exemption under which banks, credit unions, and their affiliates could claim full compliance if they were subject to a state or federal prudential regulator applying requirements substantially equivalent to the statute’s.[3] SB 189 “entirely eliminates that exemption.”[3]

    Nor does the replacement contain an entity- or data-level exemption tied to the Gramm-Leach-Bliley Act, the federal privacy statute that governs financial institutions and frequently shields them from state privacy laws.[3] The practical consequence: a bank doing business with Colorado consumers that uses scoring or ranking technology in a credit decision is now directly regulated by state AI law for the first time. As one law firm analysis put it, the unambiguous pull of regulated entities into scope “marks a new era in state AI laws for financial institutions.”[3]

    The law does carve out functions regulators did not want to burden: technologies used for anti-money-laundering and counter-terrorist-financing controls, sanctions compliance, and fraud prevention, including identity verification.[3] That split reflects a judgment about where automation is protective rather than consequential — screening for fraud and money laundering affects whether a transaction is permitted, not whether a consumer gets credit on fair terms. The line the statute draws runs between the security perimeter and the credit decision itself.


    The Borrowed Notice

    The statute’s key accommodation for lenders is a safe harbor that converts their existing compliance machinery into state compliance:

    “A financial institution that is required to provide, and does provide, a notice to a consumer under the Equal Credit Opportunity Act (ECOA) and its implementing Regulation B, and where applicable under the Fair Credit Reporting Act (FCRA), complies with the notice and disclosure requirements of SB 189 for the same decision or adverse outcome.”[3]

    The core logic is de-duplication. Federal law has required creditors to send adverse-action notices stating specific reasons for denials since 1974 — an obligation that, as the Center for Democracy and Technology (a digital-rights advocacy organization) summarizes the statutory baseline, applies “even when AI was used in any aspect of the credit decision, regardless of the AI’s complexity or opacity.”[4] Rather than layering a second, state-specific notice on top, Colorado treats the federal notice as sufficient for the same decision.[3] For banks, the incremental compliance work therefore concentrates in what federal law never required: the data-correction and human-review workflows triggered after an adverse outcome.

    The timing makes the trade stark. On April 22, 2026, the CFPB finalized a rule amending Regulation B to remove disparate-impact liability from federal credit law.[4] With effects-based enforcement gone at the federal level and the reasonable-care anti-bias duty gone from Colorado’s statute, what remains protecting a consumer denied credit by an algorithm in Colorado is the notice explaining why, the right to correct bad inputs, and the right to a human second look.[3][4] The state’s bet is that disclosure plus human review can do the work that upstream bias testing used to do — a proposition no one has yet tested.


    Today’s Hearing, and What the Rules Would Require

    The attorney general’s rulemaking has moved in parallel. Proposed rules (4 C.C.R. 904-6) were filed with the Secretary of State on August 11, 2026; an interim draft followed on October 6; and a regulatory analysis was released October 21.[5] The formal rulemaking hearing — in person at 1300 Broadway in Denver and virtually — is scheduled for October 26, 2026, the same day the written comment period closes at 11:59 p.m. Mountain Time.[5]

    The draft rules would translate the statute’s general duties into operational deadlines:[6]

    • Notice design. Post-adverse-outcome notices would go out through two channels where feasible, including an online method, and must explain the decision, the ADMT’s role, the human involvement, the principal reasons, relevant scores or inferences, and how to exercise rights.[6]
    • Response deadlines. Instructions for submitting rights requests within 24 hours of a request; responses to data and correction requests within 45 days; acknowledgment of human-review requests within 10 days; completion of human review within 45 days.[6]
    • Supply-chain documentation. A new “midstream developer” role would have to obtain upstream documentation and pass it through the AI supply chain to deployers.[6]

    The most consequential drafting choice is what the rules say a “human” review must be. The reviewer cannot be the original decision-maker, should sit outside the decision-maker’s reporting chain where feasible, must have subject-matter knowledge and the authority to approve, modify, or override the decision, and must be protected from retaliation — and, critically, “ADMT may not assist in the human review itself.”[6] That last clause is the difference between a genuine second look and a rubber stamp run past the same model. The rules would also presume review to be commercially reasonable when the adverse outcome is a severe and irreversible denial of a basic human need, subject to a demanding rebuttal.[6]


    The Shadow Over January 1

    Whether the law takes effect on schedule depends on litigation that predates the replacement. On April 9, 2026, xAI — now part of SpaceX — sued in the U.S. District Court for the District of Colorado to enjoin the 2024 law, raising First Amendment, Commerce Clause, vagueness, and equal-protection claims; the Department of Justice intervened on April 24, citing an executive order that singled out the Colorado law as problematic state AI regulation.[6] On April 27, the court entered a stipulated order under which the attorney general agreed not to enforce or investigate the 2024 statute while the case pends.[6]

    The same order preserves the challenge’s reach into the new law: xAI may move for a preliminary injunction within 28 days after final adoption of rules implementing the replacement.[6] That means the hearing that closes today starts a 28-day clock toward a possible federal injunction against a law that has not yet taken effect — an unusual posture in which compliance planning and constitutional litigation run on the same calendar, and lenders cannot prudently wait for either to resolve.[6]


    What to Watch

    • The “materially influence” gap. The draft rules declined to supply the presumptions and safe harbors stakeholders requested, leaving common screening and ranking tools to fact-specific analysis — the question that determines how many systems fall in scope at all.[6]
    • Whether xAI moves to enjoin. The 28-day window after final rule adoption is the last clear checkpoint before the January 1, 2027 effective date.[6]
    • The first enforcement signals. The attorney general must report annually on enforcement, and the 60-day cure period gives banks a preview of what the office considers a violation before penalties attach.[1]
    • The ECOA safe harbor in practice. Whether Colorado treats an adequate federal adverse-action notice as fully dispositive — or as a floor that the human-review deadlines sit on top of — will decide how much net-new work the law creates for lenders.[3]

    References

    [1] [PRIMARY] Colorado General Assembly. “SB26-189: Automated Decision-Making Technology.” Signed May 14, 2026. leg.colorado.gov

    [2] [SECONDARY] Schwartz, Tessa, Marian A. Waldmann Agarwal, and Joshua R. Fattal. “Colorado Hits Reset on AI Regulation With a New AI Act: What Developers and Deployers Need to Know.” Morrison Foerster. May 15, 2026. mofo.com

    [3] [SECONDARY] “The New Colorado AI Act: What Financial Institutions Need to Know.” Cooley Finsights. May 2026. finsights.cooley.com

    [4] [ADVOCACY] Center for Democracy and Technology. “AI in Financial Services: Current Legal Landscape.” 2026. cdt.org

    [5] [PRIMARY] Colorado Attorney General. “Colorado Automated Decision-Making Technology & Chatbot Safety Rulemaking.” Accessed October 26, 2026. coag.gov

    [6] [SECONDARY] McGinnis, Karin M., and Tiffany E. Payne. “Colorado’s Proposed ADMT Rules: What Businesses Need to Know Now.” Moore & Van Allen. October 5, 2026. mvalaw.com


    AI automated and generated content. Human reviewed.


    Listen to the Podcast: AI Policy for the Bears

    Episode 39: The Borrowed Notice: How a Six-Week Repeal, a Vanished Bank Exemption, and a Rulemaking Hearing on Its Final Comment Day Made Colorado the First State AI Law to Bind Banks Directly — click to listen

    Two bears break down this story on AI Policy for the Bears, Episode 39.

  • The Effects Test Repeal: How One July Rule, a Courtroom Filing, and a State-Law Backstop Are Redrawing AI Lending Oversight

    The Effects Test Repeal: How One July Rule, a Courtroom Filing, and a State-Law Backstop Are Redrawing AI Lending Oversight

    The CFPB stripped disparate impact from federal credit law while AI underwriting spread — and the response arrived from state capitols, state examiners, and a D.C. courtroom where the opening brief landed this week.


    One Rule Changed What “Discrimination” Means in Credit

    On April 22, 2026, the Consumer Financial Protection Bureau published a final rule rewriting Regulation B, the regulation that implements the Equal Credit Opportunity Act (ECOA). The rule, effective July 21, 2026, expressly states that ECOA does not provide for disparate impact liability and removes the “effects test” from the regulation and its official commentary[1]. It also narrowed the prohibition on “discouragement” toward a standard focused on intentional, exclusionary statements, and restricted special purpose credit programs run by for-profit lenders[1].

    The core logic of the effects test was statistical accountability: it allowed regulators and private plaintiffs to challenge a facially neutral credit practice that disproportionately harmed consumers on a protected basis — even if the harm was unintentional — unless the practice served a legitimate business need that could not be met with a less discriminatory alternative[2]. For algorithmic underwriting, this mattered enormously. A machine learning model rarely “intends” anything, but it can still reproduce and amplify the patterns embedded in its training data, and the effects test was the legal instrument built for exactly that problem. Eliminating it removes the primary lens through which algorithmic credit models were examined at the federal level. Disparate treatment liability survives, including claims based on facially neutral criteria used as proxies for protected characteristics[2] — but proving proxy use requires showing the lender chose a variable for its correlation with a protected trait, a far higher evidentiary bar than showing a measured disparity.


    The AI Underwriting Collision

    The rule lands at the exact moment AI models have become load-bearing in credit decisions. Lenders now use machine learning to score previously unscorable consumers, to set terms, and to screen applicants before they ever reach a human underwriter. Fair lending compliance for these systems has traditionally run through disparate impact testing: remove protected-class proxies, test the model’s outputs for disparities, and search for less discriminatory alternatives[3]. That discipline is what the CFPB’s rule abandons at the federal level.

    The Bureau itself has already stopped using disparate impact in supervision and enforcement, telling Congress in its spring 2026 Semi-Annual Report that it is focused instead on “uncovering direct evidence of intentional discrimination on a prohibited basis”[2]. In practice, this means the CFPB will look for smoking guns rather than statistical signatures — a posture that fits human prejudice but poorly matches how algorithmic bias actually behaves. When a model weights a shopping-habit proxy or a device type in ways that track race, there is rarely a memo saying so.


    The Courtroom Fight: The Opening Brief Is In

    The rule is already under legal challenge, and this week the case crossed its first threshold. The National Fair Housing Alliance, Rise Economy, and two fair lending analytics firms sued the CFPB in the U.S. District Court for the District of Columbia on May 27, 2026, arguing the Bureau exceeded its statutory authority and violated the Administrative Procedure Act[4]. The case — National Fair Housing Alliance et al. v. CFPB et al., No. 1:26-cv-01820, before Senior Judge Beryl Howell — did not seek a preliminary injunction, so the rule took effect on schedule. On September 29, 2026, the plaintiffs filed their motion for summary judgment on the deadline set by the court’s scheduling order[5]. The CFPB’s opposition and cross-motion are due November 10, replies run through January 12, 2027, and an appeal to the D.C. Circuit is likely regardless of outcome[4].

    The central issue is whether ECOA permits disparate impact liability at all. The CFPB concluded it does not; the plaintiffs argue the statute itself authorizes the theory, making the agency’s removal unlawful[4]. Both sides anticipate the question eventually reaching the Supreme Court[4]. In the meantime, the practical consequence is unambiguous: the revised Regulation B is in effect, and creditors are subject to it[4].

    The stakes, as the plaintiffs frame them, are the credit models themselves. The National Fair Housing Alliance describes the rule as one that “reverses 50 years of fair lending protections” and warned that eliminating disparate impact “gives lenders the green light to use credit models that unnecessarily exclude creditworthy people, without any requirement to implement fairer alternatives” (advocacy organization)[6]. That argument is aimed squarely at algorithmic underwriting: if no one must test a model for discriminatory effects, the models that most need testing are the ones least likely to be tested.


    Where Algorithmic Lending Risk Actually Lives Now

    Federal retrenchment did not delete the risk — it relocated it. Three channels now carry the exposure that the CFPB vacated:

    • Private litigation. ECOA carries a five-year statute of limitations, and both ECOA and the Fair Housing Act create private rights of action. Attorneys can review a lender’s public HMDA data, identify statistical disparities through analytics software, and file claims at a relatively low initial threshold[7]. The Sixth Circuit’s Golden v. City of Columbus assumed disparate impact claims are available under ECOA, giving private plaintiffs a doctrinal hook even as the agency retreats[2]. For an AI lender, the math is simple: the disparities a regulator might have flagged in an exam now surface as complaints instead.
    • Mortgage and housing credit. The Fair Housing Act still recognizes disparate impact claims, as the Supreme Court affirmed in Texas Department of Housing and Community Affairs v. Inclusive Communities Project (2015)[2]. Any lender extending credit secured by residential real estate — a category that includes most AI-underwritten mortgage products — remains exposed to effects-based claims through a different statute. A lender can drop its ECOA testing program and still be sued under the FHA for the same model.
    • State enforcement. States that codified disparate impact independently enforce it through their own attorneys general, and the Supreme Court’s Inclusive Communities ruling grounded the doctrine in the FHA’s statutory text rather than in agency regulations — so federal retrenchment does not touch the state statutes[7].

    There is also a temporal logic that exam-schedule changes compound. The FDIC moved to a schedule partly based on asset size and compliance rating — a well-rated bank under $350 million can now go roughly six years between joint compliance and CRA exams[7]. A bias problem in a model that goes unmonitored for six years does not stay small; it compounds across every cohort the model scores. The gap between federal check-ins is filled — or not filled — by the lender’s own monitoring.


    The State Backstop Is Building Itself

    Even as the federal effects test came out, states were writing their own in. The pattern is not random — it is the mirror image of the federal pullback, and each instrument targets a different layer of the AI lending stack:

    • New Jersey codified disparate impact. In December 2025, its Division on Civil Rights adopted the most comprehensive state-level disparate impact regulations in the country, covering employment, housing, places of public accommodation, financial lending, and contracting under the Law Against Discrimination[8], with guidance on how that liability applies to AI and automated decision-making tools[7].

    “As the Trump Administration continues its unlawful attempts to dismantle disparate impact protections at the federal level, it’s more important than ever that states take action to protect the civil rights of their residents — and that’s exactly what we’re doing.” — New Jersey Attorney General Matthew J. Platkin, December 17, 2025 (state government official)[8]

    • Connecticut put AI on notice. Its attorney general issued a February 2026 memorandum confirming that state anti-discrimination and credit laws apply to AI-driven lending decisions exactly as they do to any other lending decision[7]. The move preempts the argument that algorithmic opacity creates a legal gray zone: under Connecticut law, there is no algorithmic exception.
    • New York broadened enforcement power. The FAIR Business Practices Act, effective February 2026, extends the state’s consumer protection law beyond deceptive practices to unfair and abusive conduct, expanding the attorney general’s enforcement authority[7]. For AI lending, “unfair” is a capacious standard — it does not require proving intent or even a protected-class disparity, only substantial consumer injury that consumers cannot reasonably avoid.
    • Colorado retooled for AI decisions. After xAI sued and the U.S. Department of Justice joined to block its first-in-the-nation AI Act on Equal Protection grounds, Colorado repealed and replaced it with SB 26-189, a narrower automated decision-making regime covering consequential decisions in lending among other domains, effective January 1, 2027[9]. The law trades the original’s impact-assessment mandates for notice and explanation obligations when automated decisions harm consumers — a shift from preventing algorithmic harm to documenting it[9].

    The logic of this patchwork is that it reintroduces, through state law, much of what the federal rule removed — but unevenly. A national AI lender now operates under a single federal standard for intent, but up to fifty different standards for effects. Compliance strategy inverts accordingly: instead of designing to the federal floor, lenders increasingly design to the strictest state where they operate[7].


    Examiners Are Learning to Ask AI Questions

    The newest layer is supervisory rather than statutory. On September 16, 2026, the Conference of State Bank Supervisors (CSBS) released an Artificial Intelligence Supervisory Framework for state examiners to use when examining state-chartered banks and state-licensed nonbank financial institutions — mortgage companies, lenders, money transmitters, and debt collectors among them[10]. CSBS was explicit that the framework creates no new legal obligations; it gives examiners a discretionary, principles-based structure for identifying how institutions use AI and applying existing supervisory resources on third-party risk, model risk, consumer protection, and fair lending[10].

    What makes the framework significant is what examiners may now request: AI policies and procedures, inventories of AI systems and use cases, governance and oversight materials, risk assessments, vendor documentation, and monitoring information[11]. It is built on the National Institute of Standards and Technology’s AI Risk Management Framework, the Cyber Risk Institute’s Financial Services AI Risk Management Framework, and the U.S. Treasury’s AI Lexicon, and each state agency decides how far to incorporate it[10]. The consequence is a de facto documentation requirement: a lender that cannot produce an inventory of where AI touches its credit decisions has a problem no statute created but any examiner can find. State regulators supervise 79% of all U.S. banks, so the framework’s reach is broader than its non-binding language suggests[10].

    Notice the design logic here. Where legislation mandates outcomes, supervision mandates legibility. The CSBS framework does not tell lenders what their models may do — it tells them they must be able to explain what their models already do. For AI underwriting, that is the practical battleground of the next several years: the inventory, the risk assessment, and the explanation of an adverse decision.


    What to Watch

    • November 10, 2026 — The CFPB’s opposition and cross-motion for summary judgment are due in NFHA v. CFPB, the government’s first full answer to the argument that ECOA’s text supports effects-based liability[4].
    • October 9, 2026 — Comments close on HUD’s separate Title VI proceeding; the related proposal to strip disparate impact from the Fair Housing Act remains unfinalized after significant opposition[7]. If HUD finalizes, the FHA channel described above erodes for mortgage credit.
    • January 12, 2027 — Summary judgment briefing completes in the ECOA case, with a decision — and an appeal — to follow[4].
    • January 1, 2027 — Colorado SB 26-189 takes effect, adding notice and explanation duties for consequential AI decisions in lending[9].

    References

    [1] [PRIMARY] Consumer Financial Protection Bureau — “Equal Credit Opportunity Act (Regulation B),” final rule, 91 FR 21620 (April 22, 2026)

    [2] [SECONDARY] Venable LLP — “CFPB Makes Significant Changes to Regulation B” by James Williams and Michaela Bevan (May 27, 2026)

    [3] [SECONDARY] Brookings Institution — “An AI fair lending policy agenda for the federal financial regulators”

    [4] [SECONDARY] Consumer Finance Monitor (Ballard Spahr) — “Lawsuit Challenging CFPB’s Revised ECOA Rule Moves Forward Under Lengthy Summary Judgment Schedule” (August 10, 2026)

    [5] [PRIMARY] U.S. District Court for the District of Columbia — Docket, National Fair Housing Alliance v. CFPB, No. 1:26-cv-01820, entry 28: plaintiffs’ motion for summary judgment filed September 29, 2026

    [6] [ADVOCACY] National Fair Housing Alliance — “National Fair Housing Alliance Sues CFPB and Acting Director Vought to Stop Rollback of Fair Lending Protections” (May 27, 2026)

    [7] [SECONDARY] Ncontracts — “Fair Lending Update 2026: Disparate Impact, State Enforcement, and Underwriting Risk” by Alesha Briley, CRCM (August 25, 2026)

    [8] [PRIMARY] New Jersey Office of the Attorney General — “AG Platkin Announces Division on Civil Rights Adopts Landmark Rules on Disparate Impact Discrimination Under New Jersey Law” (December 17, 2025)

    [9] [SECONDARY] Consumer Finance Monitor (Ballard Spahr) — “Colorado rewrites its landmark AI law: Unpacking SB 26-189 and what it means for businesses” (May 12, 2026)

    [10] [PRIMARY] Conference of State Bank Supervisors — “CSBS Announces AI Supervisory Framework” (September 16, 2026)

    [11] [SECONDARY] Venable LLP — “AI Comes to State Financial Examinations: The New CSBS Supervisory Framework” by Jonathan L. Pompan (September 29, 2026)


    AI automated and generated content. Human reviewed.


    Listen to the Bears Discuss This


    Episode 31: The Effects Test Repeal: How One July Rule, a Courtroom Filing, and a State-Law Backstop Are Redrawing AI Lending Oversight — click to listen

    Episode 31 of AI Policy for the Bears