{"id":"https://openalex.org/W4412888254","doi":"https://doi.org/10.18653/v1/2025.findings-acl.713","title":"Be Cautious When Merging Unfamiliar LLMs: A Phishing Model Capable of Stealing Privacy","display_name":"Be Cautious When Merging Unfamiliar LLMs: A Phishing Model Capable of Stealing Privacy","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888254","doi":"https://doi.org/10.18653/v1/2025.findings-acl.713"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.713","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.713","pdf_url":"https://aclanthology.org/2025.findings-acl.713.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-acl.713.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5119181384","display_name":"Guo Zhenyuan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo Zhenyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012604520","display_name":"Yi Shi","orcid":"https://orcid.org/0000-0003-3575-4948"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yi Shi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032995108","display_name":"Wenlong Meng","orcid":"https://orcid.org/0009-0005-2613-0676"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wenlong Meng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045305534","display_name":"Chen Gong","orcid":"https://orcid.org/0000-0001-6178-4118"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen Gong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006624972","display_name":"Chengkun Wei","orcid":"https://orcid.org/0000-0001-8849-8808"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chengkun Wei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5101562847","display_name":"Wenzhi Chen","orcid":"https://orcid.org/0000-0003-1674-4701"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wenzhi Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"13852","last_page":"13871"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9922999739646912,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9922999739646912,"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/T11045","display_name":"Privacy, Security, and Data Protection","score":0.9868999719619751,"subfield":{"id":"https://openalex.org/subfields/3312","display_name":"Sociology and Political Science"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10237","display_name":"Cryptography and Data Security","score":0.9689000248908997,"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/phishing","display_name":"Phishing","score":0.7928353548049927},{"id":"https://openalex.org/keywords/internet-privacy","display_name":"Internet privacy","score":0.686237633228302},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6755382418632507},{"id":"https://openalex.org/keywords/computer-security","display_name":"Computer security","score":0.5885834097862244},{"id":"https://openalex.org/keywords/world-wide-web","display_name":"World Wide Web","score":0.3426628112792969},{"id":"https://openalex.org/keywords/the-internet","display_name":"The Internet","score":0.2589402198791504}],"concepts":[{"id":"https://openalex.org/C83860907","wikidata":"https://www.wikidata.org/wiki/Q135005","display_name":"Phishing","level":3,"score":0.7928353548049927},{"id":"https://openalex.org/C108827166","wikidata":"https://www.wikidata.org/wiki/Q175975","display_name":"Internet privacy","level":1,"score":0.686237633228302},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6755382418632507},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.5885834097862244},{"id":"https://openalex.org/C136764020","wikidata":"https://www.wikidata.org/wiki/Q466","display_name":"World Wide Web","level":1,"score":0.3426628112792969},{"id":"https://openalex.org/C110875604","wikidata":"https://www.wikidata.org/wiki/Q75","display_name":"The Internet","level":2,"score":0.2589402198791504}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.713","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.713","pdf_url":"https://aclanthology.org/2025.findings-acl.713.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.713","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.713","pdf_url":"https://aclanthology.org/2025.findings-acl.713.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.5099999904632568,"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10"}],"awards":[{"id":"https://openalex.org/G4929171329","display_name":null,"funder_award_id":"62302441","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320322927","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888254.pdf","grobid_xml":"https://content.openalex.org/works/W4412888254.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2149202530","https://openalex.org/W2807822918","https://openalex.org/W2921723332","https://openalex.org/W4391093354","https://openalex.org/W2482950156","https://openalex.org/W4396966040","https://openalex.org/W2305322260"],"abstract_inverted_index":{"Model":[0],"merging":[1,114,148],"is":[2],"a":[3,17,75,81,89,97,104,149,182],"widespread":[4],"technology":[5],"in":[6,48,68],"large":[7],"language":[8],"models":[9],"(LLMs)":[10],"that":[11,79,102,147],"integrates":[12],"multiple":[13],"task-specific":[14],"LLMs":[15,33,66],"into":[16,113],"unified":[18],"one,":[19],"enabling":[20],"the":[21,26,62,69,115,120,123,138,142,153,159,177],"merged":[22,139],"model":[23,49,59,70,83,99,140,151],"to":[24,107,158],"inherit":[25],"specialized":[27,105],"capabilities":[28],"of":[29,64,85,155,179],"these":[30],"LLMs.Most":[31],"taskspecific":[32],"are":[34],"sourced":[35],"from":[36],"open-source":[37],"communities":[38],"and":[39,168],"have":[40],"not":[41],"undergone":[42],"rigorous":[43],"auditing,":[44],"potentially":[45],"imposing":[46],"risks":[47],"merging.This":[50],"paper":[51],"highlights":[52],"an":[53,57],"overlooked":[54],"privacy":[55,63,76,87,91,156],"risk:":[56],"unsafe":[58],"could":[60],"compromise":[61],"other":[65],"involved":[67],"merging.Specifically,":[71],"we":[72,95],"propose":[73],"PHIMM,":[74],"attack":[77,109],"approach":[78],"trains":[80],"phishing":[82,92,116,121,143,150],"capable":[84],"stealing":[86],"using":[88],"crafted":[90],"instruction":[93],"dataset.Furthermore,":[94],"introduce":[96],"novel":[98],"cloaking":[100],"method":[101],"mimics":[103],"capability":[106],"conceal":[108],"intent,":[110],"luring":[111],"users":[112],"model.Once":[117],"victims":[118],"merge":[119],"model,":[122],"attacker":[124],"can":[125],"extract":[126],"personally":[127],"identifiable":[128],"information":[129,134],"(PII)":[130],"or":[131],"infer":[132],"membership":[133],"(MI)":[135],"by":[136,166,172],"querying":[137],"with":[141],"instruction.Experimental":[144],"results":[145,160],"show":[146],"increases":[152],"risk":[154],"breaches.Compared":[157],"before":[161],"merging,":[162],"PII":[163],"leakage":[164,170],"increased":[165,171],"3.9%":[167],"MI":[169],"17.4%":[173],"on":[174],"average.We":[175],"release":[176],"code":[178],"PHIMM":[180],"through":[181],"link.":[183]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}