{"id":"https://openalex.org/W7225086465","doi":"https://doi.org/10.3390/fintech5040101","title":"Benchmarking Security and Reliability of Open-Source and Commercial Large Language Models in Financial-Domain Scenarios","display_name":"Benchmarking Security and Reliability of Open-Source and Commercial Large Language Models in Financial-Domain Scenarios","publication_year":2026,"publication_date":"2026-11-02","ids":{"openalex":"https://openalex.org/W7225086465","doi":"https://doi.org/10.3390/fintech5040101"},"language":"en","primary_location":{"id":"doi:10.3390/fintech5040101","doi":"https://doi.org/10.3390/fintech5040101","is_oa":true,"landing_page_url":"https://doi.org/10.3390/fintech5040101","pdf_url":null,"source":{"id":"https://openalex.org/S4210236419","display_name":"FinTech","issn_l":"2674-1032","issn":["2674-1032"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["anvur-scientific","cwts-core","doaj"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"book series"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"FinTech","raw_type":"journal-article"},"volume":null,"type":"article","is_traditional":true,"indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.3390/fintech5040101","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5001761815","display_name":"Marko Pribisali\u0107","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I154347574","display_name":"University of Rijeka","ror":"https://ror.org/05r8dqr10","country_code":"HR","type":"education","lineage":["https://openalex.org/I154347574"]}],"countries":["HR"],"is_corresponding":true,"raw_author_name":"Marko Pribisali\u0107","raw_affiliation_strings":["Faculty of Informatics and Digital Technologies, University of Rijeka, Radmile Matej\u010di\u0107 2, 51000 Rijeka, Croatia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Informatics and Digital Technologies, University of Rijeka, Radmile Matej\u010di\u0107 2, 51000 Rijeka, Croatia","institution_ids":["https://openalex.org/I154347574"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5040727386","display_name":"Sanda Martin\u010di\u0107-Ip\u0161i\u0107","orcid":"https://orcid.org/0000-0002-1900-5333","observed_orcids":["https://orcid.org/0000-0002-1900-5333"]},"institutions":[{"id":"https://openalex.org/I154347574","display_name":"University of Rijeka","ror":"https://ror.org/05r8dqr10","country_code":"HR","type":"education","lineage":["https://openalex.org/I154347574"]}],"countries":["HR"],"is_corresponding":false,"raw_author_name":"Sanda Martin\u010di\u0107-Ip\u0161i\u0107","raw_affiliation_strings":["Center for Artificial Intelligence and Cybersecurity, University of Rijeka, Radmile Matej\u010di\u0107 2, 51000 Rijeka, Croatia","Faculty of Informatics and Digital Technologies, University of Rijeka, Radmile Matej\u010di\u0107 2, 51000 Rijeka, Croatia"],"raw_orcid":"https://orcid.org/0000-0002-1900-5333","affiliations":[{"raw_affiliation_string":"Center for Artificial Intelligence and Cybersecurity, University of Rijeka, Radmile Matej\u010di\u0107 2, 51000 Rijeka, Croatia","institution_ids":["https://openalex.org/I154347574"]},{"raw_affiliation_string":"Faculty of Informatics and Digital Technologies, University of Rijeka, Radmile Matej\u010di\u0107 2, 51000 Rijeka, Croatia","institution_ids":["https://openalex.org/I154347574"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5001761815"],"corresponding_institution_ids":["https://openalex.org/I154347574"],"apc_list":{"value":1000,"currency":"CHF","value_usd":1207},"apc_paid":{"value":1000,"currency":"CHF","value_usd":1207},"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.82378635,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"5","issue":"4","first_page":"101","last_page":"101"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9437999725341797,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9437999725341797,"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/T10028","display_name":"Topic Modeling","score":0.0478999987244606,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.0020000000949949026,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9860000014305115},{"id":"https://openalex.org/keywords/llm-benchmarking","display_name":"LLM benchmarking","score":0.8019999861717224},{"id":"https://openalex.org/keywords/regulatory-compliance","display_name":"regulatory compliance","score":0.9200000166893005},{"id":"https://openalex.org/keywords/adversarial-robustness","display_name":"adversarial robustness","score":0.9769999980926514},{"id":"https://openalex.org/keywords/financial-services","display_name":"financial services","score":0.4860000014305115},{"id":"https://openalex.org/keywords/llm-evaluation-framework","display_name":"LLM evaluation framework","score":0.7239999771118164},{"id":"https://openalex.org/keywords/model-selection","display_name":"model selection","score":0.36800000071525574}],"concepts":[{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.8037999868392944},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5928999781608582},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.5695000290870667},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.4410000145435333},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.43970000743865967},{"id":"https://openalex.org/C93959086","wikidata":"https://www.wikidata.org/wiki/Q6888345","display_name":"Model selection","level":2,"score":0.29249998927116394},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.28850001096725464},{"id":"https://openalex.org/C121822524","wikidata":"https://www.wikidata.org/wiki/Q5157582","display_name":"Computer security model","level":2,"score":0.28630000352859497},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.2799000144004822},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.26829999685287476},{"id":"https://openalex.org/C38369872","wikidata":"https://www.wikidata.org/wiki/Q7445009","display_name":"Security analysis","level":2,"score":0.2667999863624573},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.26260000467300415},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.2603999972343445},{"id":"https://openalex.org/C139945424","wikidata":"https://www.wikidata.org/wiki/Q1940696","display_name":"Mean squared error","level":2,"score":0.2522999942302704}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.3390/fintech5040101","doi":"https://doi.org/10.3390/fintech5040101","is_oa":true,"landing_page_url":"https://doi.org/10.3390/fintech5040101","pdf_url":null,"source":{"id":"https://openalex.org/S4210236419","display_name":"FinTech","issn_l":"2674-1032","issn":["2674-1032"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["anvur-scientific","cwts-core","doaj"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"book series"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"FinTech","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.3390/fintech5040101","doi":"https://doi.org/10.3390/fintech5040101","is_oa":true,"landing_page_url":"https://doi.org/10.3390/fintech5040101","pdf_url":null,"source":{"id":"https://openalex.org/S4210236419","display_name":"FinTech","issn_l":"2674-1032","issn":["2674-1032"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["anvur-scientific","cwts-core","doaj"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"book series"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"FinTech","raw_type":"journal-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":65,"referenced_works":["https://openalex.org/W1546425147","https://openalex.org/W1825242443","https://openalex.org/W2053154970","https://openalex.org/W2078483536","https://openalex.org/W2084413241","https://openalex.org/W2118978333","https://openalex.org/W2141766660","https://openalex.org/W2164777277","https://openalex.org/W2535690855","https://openalex.org/W2767106145","https://openalex.org/W2798658104","https://openalex.org/W2979826702","https://openalex.org/W3021916629","https://openalex.org/W3045321166","https://openalex.org/W3201174429","https://openalex.org/W4252684946","https://openalex.org/W4299627282","https://openalex.org/W4317748910","https://openalex.org/W4362515116","https://openalex.org/W4378189609","https://openalex.org/W4382998379","https://openalex.org/W4385565472","https://openalex.org/W4385573185","https://openalex.org/W4385573863","https://openalex.org/W4385894687","https://openalex.org/W4387321091","https://openalex.org/W4388886073","https://openalex.org/W4389403907","https://openalex.org/W4389518784","https://openalex.org/W4389519254","https://openalex.org/W4390231192","https://openalex.org/W4392773006","https://openalex.org/W4396962699","https://openalex.org/W4399750638","https://openalex.org/W4400679025","https://openalex.org/W4401042900","https://openalex.org/W4402057292","https://openalex.org/W4402665833","https://openalex.org/W4402667035","https://openalex.org/W4402683773","https://openalex.org/W4402685935","https://openalex.org/W4402692877","https://openalex.org/W4402858027","https://openalex.org/W4403499178","https://openalex.org/W4404782181","https://openalex.org/W4404782283","https://openalex.org/W4404783401","https://openalex.org/W4405310541","https://openalex.org/W4406264456","https://openalex.org/W4411120158","https://openalex.org/W4412886673","https://openalex.org/W4412887937","https://openalex.org/W4412888616","https://openalex.org/W4415109192","https://openalex.org/W4415796224","https://openalex.org/W4415797606","https://openalex.org/W4415798415","https://openalex.org/W7131794385","https://openalex.org/W7133189643","https://openalex.org/W7133204634","https://openalex.org/W7133224126","https://openalex.org/W7133235460","https://openalex.org/W7133239970","https://openalex.org/W7147374400","https://openalex.org/W7155509515"],"related_works":["https://openalex.org/W7166838916","https://openalex.org/W4386081064","https://openalex.org/W7084027548","https://openalex.org/W7131418935","https://openalex.org/W4415796224","https://openalex.org/W7221723131","https://openalex.org/W6891859504","https://openalex.org/W4417291462","https://openalex.org/W4414991998","https://openalex.org/W6910010979"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"are":[4],"adopted":[5],"in":[6,38,195],"financial":[7,40,165,197],"services,":[8],"yet":[9],"existing":[10],"benchmarks":[11,57],"provide":[12],"limited":[13],"insight":[14],"into":[15],"financial-domain":[16,56],"capability,":[17],"regulatory":[18,67],"compliance-action":[19,68],"performance,":[20],"and":[21,35,70,83,97,112,144,167,183,190],"security":[22,115,147,168],"robustness.":[23,77],"This":[24],"study":[25],"proposes":[26],"a":[27,98,181],"unified":[28],"multidimensional":[29,187],"evaluation":[30,156,189],"framework":[31,43,179],"for":[32,58,65,75,186],"evaluating":[33],"open-source":[34,82,127,141],"commercial":[36,85],"LLMs":[37],"regulated":[39,196],"environments.":[41,198],"The":[42,177],"comprises":[44],"three":[45,84],"complementary":[46],"benchmark":[47],"components:":[48],"the":[49,61,71,107,121,139],"Integrated":[50],"Financial":[51],"Benchmark":[52],"(IFB),":[53],"which":[54],"integrates":[55],"capability":[59,166],"evaluation;":[60],"newly":[62,72],"developed":[63,73],"RegBench":[64,123,142],"context-grounded":[66],"performance;":[69],"FinSec-Eval-1000":[74],"adversarial":[76],"Eight":[78],"LLMs,":[79],"including":[80],"five":[81],"models,":[86,128],"were":[87],"evaluated":[88],"using":[89],"an":[90],"inference":[91],"pipeline,":[92],"task-specific":[93],"metrics,":[94],"statistical":[95],"testing,":[96],"Composite":[99],"Security":[100],"Score":[101],"(SS).":[102],"Claude":[103],"Sonnet":[104],"4.6":[105],"achieved":[106,120,138],"highest":[108,122],"IFB":[109,134],"performance":[110,124],"(68.72%)":[111],"four-judge":[113,145],"mean":[114,146],"robustness":[116,169],"(97.43%),":[117],"while":[118],"GPT-5.1":[119],"(50.28%).":[125],"Among":[126],"Gemma":[129],"3":[130],"12B":[131],"led":[132],"on":[133],"(63.50%),":[135],"whereas":[136],"Phi-3-small":[137],"strongest":[140],"(49.22%)":[143],"(92.78%)":[148],"results,":[149],"demonstrating":[150],"that":[151],"model":[152,193],"rankings":[153],"vary":[154],"across":[155],"dimensions.":[157],"No":[158],"detectable":[159],"monotonic":[160],"association":[161],"was":[162],"observed":[163],"between":[164],"(Spearman\u2019s":[170],"\u03c1":[171],"=":[172,175],"0.095,":[173],"p":[174],"0.823).":[176],"benchmarking":[178],"provides":[180],"transparent":[182],"reproducible":[184],"methodology":[185],"LLM":[188],"supports":[191],"informed":[192],"selection":[194]},"counts_by_year":[],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2026-11-03T00:00:00"}