{"id":"https://openalex.org/W4416198676","doi":"https://doi.org/10.1145/3768292.3770355","title":"Unmasking Bias in Financial AI: A Robust Framework for Evaluating and Mitigating Hidden Biases in LLMs","display_name":"Unmasking Bias in Financial AI: A Robust Framework for Evaluating and Mitigating Hidden Biases in LLMs","publication_year":2025,"publication_date":"2025-11-14","ids":{"openalex":"https://openalex.org/W4416198676","doi":"https://doi.org/10.1145/3768292.3770355"},"language":null,"primary_location":{"id":"doi:10.1145/3768292.3770355","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3768292.3770355","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 6th ACM International Conference on AI in Finance","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120350756","display_name":"Shreshth Mehrotra","orcid":"https://orcid.org/0009-0000-6980-2392"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Shreshth Mehrotra","raw_affiliation_strings":["Mastercard, Gurugram, India"],"raw_orcid":"https://orcid.org/0009-0000-6980-2392","affiliations":[{"raw_affiliation_string":"Mastercard, Gurugram, India","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120547542","display_name":"P Raghavendra","orcid":"https://orcid.org/0009-0009-1767-8276"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Raghavendra P","raw_affiliation_strings":["Mastercard, Gurugram, India"],"raw_orcid":"https://orcid.org/0009-0009-1767-8276","affiliations":[{"raw_affiliation_string":"Mastercard, Gurugram, India","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120447296","display_name":"Balraj Prajesh","orcid":"https://orcid.org/0009-0006-9014-2871"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Balraj Prajesh","raw_affiliation_strings":["Mastercard, Gurugram, India"],"raw_orcid":"https://orcid.org/0009-0006-9014-2871","affiliations":[{"raw_affiliation_string":"Mastercard, Gurugram, India","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120597301","display_name":"Hrishikesh Kambale","orcid":"https://orcid.org/0009-0006-8600-5794"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hrishikesh Kambale","raw_affiliation_strings":["Mastercard, Gurugram, India"],"raw_orcid":"https://orcid.org/0009-0006-8600-5794","affiliations":[{"raw_affiliation_string":"Mastercard, Gurugram, India","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5009237168","display_name":"Puspita Majumdar","orcid":"https://orcid.org/0000-0002-4031-785X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Puspita Majumdar","raw_affiliation_strings":["Mastercard, Gurugram, India"],"raw_orcid":"https://orcid.org/0000-0002-4031-785X","affiliations":[{"raw_affiliation_string":"Mastercard, Gurugram, India","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5120350756"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.1733,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.91231012,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"951","last_page":"959"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.35019999742507935,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.35019999742507935,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.09000000357627869,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.08320000022649765,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.5629000067710876},{"id":"https://openalex.org/keywords/behavioral-economics","display_name":"Behavioral economics","score":0.3790000081062317},{"id":"https://openalex.org/keywords/financial-market","display_name":"Financial market","score":0.36820000410079956},{"id":"https://openalex.org/keywords/foundation","display_name":"Foundation (evidence)","score":0.3375999927520752},{"id":"https://openalex.org/keywords/public-finance","display_name":"Public finance","score":0.26989999413490295}],"concepts":[{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.5629000067710876},{"id":"https://openalex.org/C10138342","wikidata":"https://www.wikidata.org/wiki/Q43015","display_name":"Finance","level":1,"score":0.5393999814987183},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4023999869823456},{"id":"https://openalex.org/C109574028","wikidata":"https://www.wikidata.org/wiki/Q647525","display_name":"Behavioral economics","level":2,"score":0.3790000081062317},{"id":"https://openalex.org/C19244329","wikidata":"https://www.wikidata.org/wiki/Q208697","display_name":"Financial market","level":2,"score":0.36820000410079956},{"id":"https://openalex.org/C162118730","wikidata":"https://www.wikidata.org/wiki/Q1128453","display_name":"Actuarial science","level":1,"score":0.35899999737739563},{"id":"https://openalex.org/C2780966255","wikidata":"https://www.wikidata.org/wiki/Q5474306","display_name":"Foundation (evidence)","level":2,"score":0.3375999927520752},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3285999894142151},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.2750999927520752},{"id":"https://openalex.org/C178283979","wikidata":"https://www.wikidata.org/wiki/Q274490","display_name":"Public finance","level":2,"score":0.26989999413490295},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.2646999955177307},{"id":"https://openalex.org/C12174686","wikidata":"https://www.wikidata.org/wiki/Q1058438","display_name":"Risk assessment","level":2,"score":0.25999999046325684}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3768292.3770355","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3768292.3770355","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 6th ACM International Conference on AI in Finance","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":14,"referenced_works":["https://openalex.org/W2893425640","https://openalex.org/W2963078909","https://openalex.org/W2971307358","https://openalex.org/W3123340107","https://openalex.org/W3134354193","https://openalex.org/W3176477796","https://openalex.org/W3207316473","https://openalex.org/W4388687124","https://openalex.org/W4399528455","https://openalex.org/W4402667124","https://openalex.org/W4402671039","https://openalex.org/W4404782667","https://openalex.org/W4410609100","https://openalex.org/W4415797457"],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4],"increasingly":[5],"used":[6],"in":[7,38,53,174],"finance":[8,59,82],"for":[9,89,166],"tasks":[10],"like":[11],"market":[12],"analysis,":[13,17],"customer":[14],"support,":[15],"sentiment":[16],"and":[18,25,36,49,56,63,84,111,140,170],"automated":[19],"reporting.":[20],"However,":[21],"LLMs":[22,136],"often":[23],"inherit":[24],"perpetuate":[26],"biases":[27],"from":[28],"their":[29],"training":[30],"data,":[31],"raising":[32],"concerns":[33],"about":[34],"fairness":[35,65],"accuracy":[37],"high-stakes":[39],"financial":[40,175],"applications.":[41,177],"While":[42],"other":[43],"domains":[44],"such":[45],"as":[46],"medicine,":[47],"law,":[48],"education":[50],"have":[51],"advanced":[52],"identifying,":[54],"measuring,":[55],"reducing":[57],"bias,":[58,148],"lacks":[60],"domain-specific":[61],"datasets":[62],"robust":[64],"metrics.":[66],"To":[67],"address":[68],"this,":[69],"we":[70,119],"introduce":[71],"the":[72,81,150,167],"FinBias":[73],"dataset":[74],"which":[75,108],"includes":[76],"bias-eliciting":[77],"prompts":[78],"related":[79],"to":[80],"domain,":[83],"a":[85,101,121,163],"comprehensive":[86],"evaluation":[87,169],"framework":[88],"publicly":[90,134],"available":[91,135],"LLMs,":[92],"including":[93],"robustness":[94],"tests":[95],"against":[96],"jailbreaking.":[97],"We":[98],"also":[99],"propose":[100],"new":[102],"metric,":[103],"SAFE":[104],"(Safety-Adjusted":[105],"Fairness":[106],"Evaluation),":[107],"penalizes":[109],"stereotypical":[110],"refusal":[112],"responses":[113],"while":[114],"rewarding":[115],"debiased":[116],"outputs.":[117],"Additionally,":[118],"present":[120],"prompt":[122,152],"engineering-based":[123,153],"mitigation":[124,154,171],"strategy":[125,155],"that":[126,143],"effectively":[127,156],"reduces":[128,157],"bias.":[129,159],"Experiments":[130],"conducted":[131],"on":[132],"three":[133],"-":[137],"Mixtral,":[138],"Gemma,":[139],"LLaMA":[141],"demonstrate":[142],"these":[144],"models":[145],"exhibit":[146],"significant":[147],"but":[149],"proposed":[151],"this":[158],"This":[160],"research":[161],"provides":[162],"practical":[164],"foundation":[165],"detection,":[168],"of":[172],"bias":[173],"LLM":[176]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-11-14T00:00:00"}