[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81921-en":3,"doc-seo-81921-105":31,"detail-sidebar-cat-0-en-105":93},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},81921,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Counterfactual Methods for Detecting Unfairness in Anti-Money Laundering Algorithms","Machine learning–based predictive systems for Anti-Money Laundering (AML) process large volumes of transaction data and often incorporate sensitive client information, creating fairness concerns that are frequently overlooked. The work applies counterfactual techniques to decompose direct and indirect effects of potentially sensitive features on model predictions. Using the synthetic IBMAMLSim transaction dataset, it adds country-related features and derived behavioral statistics to improve accuracy. Counterfactual, path-specific effect analysis links greater fairness violations to the models that most benefit from the extended features.","arXiv :2607 .05 10 1v 1 [ cs .LG] 6 Jul 2026  \nCounterfactual Methods for Detecting Unfairness in Anti-Money Laundering Algorithms  \nLEA MULTERER, IDSIA, Switzerland MICHELE INCHINGOLO, IDSIA, Switzerland DAVID KLETZ, IDSIA, Switzerland  \nADRIAN COSMA, IDSIA, Switzerland ALESSANDRO ANTONUCCI, IDSIA, Switzerland  \nMARTINA GOGOVA, UBS Switzerland AG and its affiliates, Germany  \nThe application of machine learning–based predictive algorithms to Anti-Money Laundering (AML) has grown rapidly, driven by the vast volume of financial transaction data available to banks. These algorithms are typically trained not only on transactional data but also on sensitive client information, which may raise fairness concerns. Despite this, AML detection systems remain largely underexplored from a fairness perspective, even though deeper analytical methods based on counterfactuals are now available. Such techniques enable the decomposition of the direct and indirect effects of potentially sensitive features on model predictions, thereby supporting the evaluation of whether their influence is acceptable from a fairness perspective. Closing this gap, we consider the synthetic IBMAMLSim transaction dataset and construct additional features of the country of an account and its average behaviour. This improves the predictive performance of diverse machine learning models, ranging from baseline decision trees to state-of-the-art graph neural networks. We assess the potential unfairness associated with these features through a counterfactual, path-specific effect analysis. This reveals that fairness violations tend to be more pronounced for models whose predictive performance benefits the most from the extended features. Such a finding highlights a concrete instance of the trade-off between predictive accuracy and fairness in AML applications, thus underscoring the urgency of a systematic fairness analysis in such critical domains.  \nKeywords: Counterfactual Fairness Analysis, Mediation Analysis, Structural Causal Models, Graph Neural Networks, AntiMoney Laundering, Know Your Customer Data.  \nReference Format:  \nLea Multerer, Michele Inchingolo, David Kletz, Adrian Cosma, Alessandro Antonucci, and Martina Gogova. 2026. Counterfactual Methods for Detecting Unfairness in Anti-Money Laundering Algorithms. In Proceedings of Fifth European Conference on Algorithmic Fairness (ECAF’26) . Proceedings of Machine Learning Research, 16 pages.  \nAuthors’ Contact Information: Lea Multerer, IDSIA, Lugano, Switzerland, [lea.multerer@idsia.ch](lea.multerer@idsia.ch); Michele Inchingolo, IDSIA, Lugano, Switzerland, michele.inchingolo@idsia.ch; David Kletz, IDSIA, Lugano, Switzerland, [david.kletz@idsia.ch](david.kletz@idsia.ch); Adrian Cosma, IDSIA, Lugano,  \nSwitzerland, [adrian.cosma@idsia.ch](adrian.cosma@idsia.ch); Alessandro Antonucci, IDSIA, Lugano, Switzerland, [alessandro.antonucci@idsia.ch](alessandro.antonucci@idsia.ch); Martina Gogova,  \nUBS Switzerland AG and its affiliates, Frankfurt, Germany, martina.gogova@ubs.com.  \nThis paper is published under the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International (CC-BY-NC-ND 4.0) license. Authors reserve their rights to disseminate the work on their personal and corporate Web sites with the appropriate attribution.  \nECAF’26, September 02–September 04, 2026, Ghent, BE © 2026 Copyright held by the owner/author(s) .  \nProceedings of ECAF’26 . September 02 – September 04, 2026 . Ghent, BE.  \n2 • Lea Multerer, Michele Inchingolo, David Kletz, Adrian Cosma, Alessandro Antonucci, and Martina Gogova  \n1 Introduction  \nAnti-Money Laundering (AML) refers to the set of practices aimed at preventing the movement of illicit funds through the financial system. The responsibility for detecting such activities primarily lies with financial institutions, which are expected to implement Know Your Customer (KYC) standards, monitor transactions, suspend accounts deemed suspicious, and submit timely reports to regulat","cbCaieq9vCcPwHpp","https://ap.wps.com/l/cbCaieq9vCcPwHpp","pdf",283719,5,1,16,"English","en",105,"# Introduction\n## Anti-Money Laundering and AML ML Pipelines\n## Supervised Learning Challenges in AML\n## KYC Features and Regulatory Context","[{\"question\":\"Why is fairness analysis important for AML detection algorithms?\",\"answer\":\"AML classifiers often use both transaction data and sensitive client information from KYC processes. The paper argues that fairness issues are underexplored despite these sensitive inputs.\"},{\"question\":\"What role do counterfactual methods play in this work?\",\"answer\":\"Counterfactual techniques are used to decompose the direct and indirect effects of potentially sensitive features on predictions. This enables evaluation of whether feature influence is acceptable from a fairness perspective.\"},{\"question\":\"How do the added country-related features affect predictive performance and fairness?\",\"answer\":\"The synthetic IBMAMLSim dataset is extended with country features and average behavior statistics, improving predictive performance across several model types. Counterfactual path-specific analysis shows fairness violations are more pronounced for the models that benefit most from these extended features.\"}]","Counterfactual Methods for Detecting Unfairness in Anti-Money Laundering Algorithms | PDF",1784177054,40,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":88,"head_meta":90,"extra_data":92,"updated_unix":29},"counterfactual-methods-for-detecting-unfairness-in-anti-money-laundering-algorithms","",{"@graph":37,"@context":87},[38,55,70],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,52],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":51},"https://docshare.wps.com/document/research-report/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/counterfactual-methods-for-detecting-unfairness-in-anti-money-laundering-algorithms/81921/",4,{"url":53,"name":13,"@type":56,"author":57,"headline":13,"publisher":59,"fileFormat":62,"inLanguage":24,"description":14,"dateModified":63,"datePublished":64,"encodingFormat":62,"isAccessibleForFree":65,"interactionStatistic":66},"DigitalDocument",{"name":9,"@type":58},"Person",{"url":42,"name":60,"@type":61},"DocShare","Organization","application/pdf","2026-07-29","2026-07-16",true,{"@type":67,"interactionType":68,"userInteractionCount":20},"InteractionCounter",{"@type":69},"ViewAction",{"@type":71,"mainEntity":72},"FAQPage",[73,79,83],{"name":74,"@type":75,"acceptedAnswer":76},"Why is fairness analysis important for AML detection algorithms?","Question",{"text":77,"@type":78},"AML classifiers often use both transaction data and sensitive client information from KYC processes. The paper argues that fairness issues are underexplored despite these sensitive inputs.","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"What role do counterfactual methods play in this work?",{"text":82,"@type":78},"Counterfactual techniques are used to decompose the direct and indirect effects of potentially sensitive features on predictions. This enables evaluation of whether feature influence is acceptable from a fairness perspective.",{"name":84,"@type":75,"acceptedAnswer":85},"How do the added country-related features affect predictive performance and fairness?",{"text":86,"@type":78},"The synthetic IBMAMLSim dataset is extended with country features and average behavior statistics, improving predictive performance across several model types. Counterfactual path-specific analysis shows fairness violations are more pronounced for the models that benefit most from these extended features.","https://schema.org",{"og:url":53,"og:type":89,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":91,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":94},[95,99,103,107,111,116,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":96,"show_sort_weight":97,"slug":98},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":101,"slug":102},"Literature",80,"literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":104,"show_sort_weight":105,"slug":106},"Exam",70,"exam",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":30,"slug":119},7,"Healthcare","healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":47,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":47,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":47,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":47,"category_name":137,"show_sort_weight":20,"slug":138},19,"General","general"]