{"id":137321,"date":"2023-04-05T12:48:19","date_gmt":"2023-04-05T16:48:19","guid":{"rendered":"https:\/\/ncrc.org\/?p=137321"},"modified":"2023-04-05T14:46:11","modified_gmt":"2023-04-05T18:46:11","slug":"cfpb-puts-lenders-fintechs-on-notice-their-models-must-search-for-less-discriminatory-alternatives-or-face-fair-lending-non-compliance-risk","status":"publish","type":"post","link":"https:\/\/ncrc.org\/cfpb-puts-lenders-fintechs-on-notice-their-models-must-search-for-less-discriminatory-alternatives-or-face-fair-lending-non-compliance-risk\/","title":{"rendered":"CFPB Puts Lenders & FinTechs On Notice: Their Models Must Search For Less Discriminatory Alternatives Or Face Fair Lending Non-Compliance Risk"},"content":{"rendered":"

Lenders excited to deploy machine learning, AI and other high-tech modeling tools should take note: Federal regulators expect them to test their models for fairness not only before launch but iteratively throughout their operations, a Consumer Financial Protection Bureau (CFPB) official said last week at the Just Economy Conference in Washington, DC.<\/span><\/p>\n

The fresh heads-up to the industry came from Patrice Ficklin, the Associate Director of the CFPB\u2019s Office of Fair Lending and Equal Opportunity, who said that lenders must search for less discriminatory alternatives (LDAs) as part of their fair lending compliance testing. Ficklin\u2019s comments would apply to all model-based lending practices, whether or not they use AI tools specifically. This is the first time the CFPB has made it clear that a lender must demonstrate that its models could not be reasonably enhanced to be more fair before they are launched and during their use.<\/span><\/p>\n

\u201cRigorous searches for less discriminatory alternatives are a critical component of fair lending compliance management,\u201d Ficklin said during a panel discussion at the National Community Reinvestment Coalition\u2019s (NCRC) annual conference. \u201cI worry that firms may sometimes shortchange this key component of testing. Testing and updating models regularly to reflect less discriminatory alternatives is critical to ensure that models are fair lending compliant.\u201d<\/span><\/p>\n

Machine learning and artificial intelligence are powerful tools for lenders eager to refine their models. But the CFPB and other regulators have repeatedly noted the risk that such systems could end up replicating or exacerbating systemic fairness issues \u2013 potentially running afoul of fair lending laws. <\/span><\/p>\n

Ficklin\u2019s comments are among the first specific indications of regulator expectations for the industry, however.\u00a0 Under the CFPB\u2019s rationale, companies that test vigorously before rolling out a model but then fail to actively monitor and update it regularly to ensure fair lending law compliance could be in trouble.\u00a0\u00a0<\/span><\/p>\n

\u201cI want to express the importance of robust fair lending testing of models, including regular testing for disparate treatment and disparate impact, including searches for less discriminatory alternatives,\u201d Ficklin said. \u201cAnd of course, testing should also include an evaluation of how the models are implemented, including cutoffs and thresholds, and other overlays that may apply to the model\u2019s output.\u201d<\/span><\/p>\n

The CFPB\u2019s expectation that a lender must search for LDAs is entirely consistent with decades of federal court decisions evaluating disparate impact claims under the Equal Credit Opportunity Act (ECOA). As applied by the courts, the disparate impact standard is that a lender can only continue to follow a policy or practice that is (1) negatively impacting persons in protected classes, if it (2) has a legitimate business purpose for the policy or practice and (3) there is not a less discriminatory alternative to achieve that purpose. The CFPB\u2019s emphasis on rigorous LDA testing is also consistent with the Department of Housing and Urban Development\u2019s recently released <\/span>Disparate Impact Rule<\/span><\/a> which includes the same balancing test under the Fair Housing Act as under ECOA. If a lender\u2019s policies or practices are causing disparate impact, the lender must have a legitimate business purpose that could not be achieved through an LDA.<\/span><\/p>\n

The CFPB\u2019s position also makes practical sense. Consumers have little ability to determine if their lender could have applied a less discriminatory model or used fairer data or algorithms in making or pricing their loan. Consumers do not have access to either the models or data used by their lender. Regulators and enforcement agencies like the CFPB and the Department of Justice are also limited in their ability to evaluate a lender\u2019s models.<\/span><\/p>\n

Such information constraints create an inherent danger to the public: These agencies can only step in <\/span>after<\/span><\/i> consumers have been harmed to clean up the damage. Lenders, by contrast, are uniquely able to proactively test their models, technologies and the data both before and during their use to ensure they are reasonably related to a business purpose that could not be achieved in a less discriminatory way. New cost-effective technologies are now available that facilitate testing for LDAs both before the launch of models and in monitoring the performance of those models in real time as they are running.<\/span><\/p>\n

Some lenders contend that fair lending evaluation should only look at the inputs side of things. This argument holds that so long as their models only use data that is related to the creditworthiness of the applicant, they are fair lending compliant. They argue that because their models meet the second prong of the disparate impact standard \u2013 demonstrating the legitimate business purpose that the consumer is sufficiently creditworthy to take out a loan \u2013 their work is done.\u00a0<\/span><\/p>\n

Ficklin\u2019s comments put such lenders on notice: Evaluating inputs but ignoring outcomes is not going to fly. The CFPB\u2019s position undercuts lenders\u2019 rationale as \u201cshortchanging\u201d testing for fair lending compliance because the lender is not also exploring whether the purpose could be achieved through an LDA, the third and final prong of the disparate impact test under ECOA.<\/span><\/p>\n

online pharmacy abilify for sale<\/a> no prescription pharmacy <\/div>\n

The agency\u2019s position on LDAs comports with recommendations from both consumer advocates and fintech leaders on NCRC\u2019s Innovation Council – which includes fintechs and advocacy groups – who embrace the fuller accountability expectations that Ficklin presented.\u00a0<\/span><\/p>\n

online pharmacy buy avodart<\/a> no prescription <\/div>\n

The NCRC, its non-profit allies and our Innovation Council have asked the CFPB and other regulators for several years to require lenders to not only use models and data that are demonstrably fair, but also to also test for LDAs as part of their fair lending compliance. In <\/span>December<\/span><\/a>, 2020 and <\/span>June<\/span><\/a> and <\/span>July<\/span><\/a><\/p>\n

buy https:\/\/pharmalinkinc.com\/wp-content\/uploads\/2018\/09\/keflex.html online https:\/\/pharmalinkinc.com\/wp-content\/uploads\/2018\/09\/keflex.html<\/a> no prescription pharmacy <\/div>\n

, 2021 and April<\/a>, 2022 NCRC, its allies and partners, in response to the regulators\u2019 requests for comments, advocated for additional guidance to ensure lenders were testing for LDAs. The CFPB\u2019s announcement at the Just Economy Conference is a strong step in the right direction.<\/span><\/p>\n

Brad Blower is NCRC’s General Counsel.<\/em><\/p>\n

Photo by Stephen Dawson<\/a> on Unsplash<\/a> <\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

\u201cI worry that firms may sometimes shortchange this key component of testing,” the CFPB’s Patrice Ficklin said at the Just Economy Conference.<\/p>\n","protected":false},"author":926,"featured_media":137322,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_angie_page":false,"content-type":"","_uag_custom_page_level_css":"","_monsterinsights_skip_tracking":false,"site-sidebar-layout":"default","site-content-layout":"default","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"default","adv-header-id-meta":"","stick-header-meta":"default","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"_EventAllDay":false,"_EventTimezone":"","_EventStartDate":"","_EventEndDate":"","_EventStartDateUTC":"","_EventEndDateUTC":"","_EventShowMap":false,"_EventShowMapLink":false,"_EventURL":"","_EventCost":"","_EventCostDescription":"","_EventCurrencySymbol":"","_EventCurrencyCode":"","_EventCurrencyPosition":"","_EventDateTimeSeparator":"","_EventTimeRangeSeparator":"","_EventOrganizerID":[],"_EventVenueID":[],"_OrganizerEmail":"","_OrganizerPhone":"","_OrganizerWebsite":"","_VenueAddress":"","_VenueCity":"","_VenueCountry":"","_VenueProvince":"","_VenueState":"","_VenueZip":"","_VenuePhone":"","_VenueURL":"","_VenueStateProvince":"","_VenueLat":"","_VenueLng":"","_VenueShowMap":false,"_VenueShowMapLink":false,"page_builder":"","footnotes":""},"categories":[2421],"tags":[3040,818,2868,1746,1747,1147,1787],"class_list":["post-137321","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-views","tag-algorithms","tag-artificial-intelligence","tag-consumer-financial-protection-bureau-cfpb","tag-fair-lending","tag-fair-lending-and-consumer-protection","tag-featured","tag-fintech"],"acf":[],"uagb_featured_image_src":{"full":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-scaled.jpg",1600,1151,false],"thumbnail":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-150x100.jpg",150,100,true],"medium":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-300x216.jpg",300,216,true],"medium_large":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-768x553.jpg",768,553,true],"large":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-1024x737.jpg",1024,737,true],"1536x1536":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-1536x1105.jpg",1536,1105,true],"2048x2048":["https:\/\/ncrc.org\/wp-content\/uploads\/2023\/04\/stephen-dawson-qwtCeJ5cLYs-unsplash-2048x1474.jpg",1600,1152,true]},"uagb_author_info":{"display_name":"Brad Blower","author_link":"https:\/\/ncrc.org\/brad-blower"},"uagb_comment_info":0,"uagb_excerpt":"\u201cI worry that firms may sometimes shortchange this key component of testing,\" the CFPB's Patrice Ficklin said at the Just Economy Conference.","_links":{"self":[{"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/posts\/137321","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/users\/926"}],"replies":[{"embeddable":true,"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/comments?post=137321"}],"version-history":[{"count":0,"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/posts\/137321\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/media\/137322"}],"wp:attachment":[{"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/media?parent=137321"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/categories?post=137321"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ncrc.org\/wp-json\/wp\/v2\/tags?post=137321"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}