{"id":"https://openalex.org/W7225167649","doi":"https://doi.org/10.48550/arxiv.2611.02689","title":"The Customer Never Sees the Goal: Measuring and Detecting Vulnerability-Targeted Manipulation in LLM Assistants","display_name":"The Customer Never Sees the Goal: Measuring and Detecting Vulnerability-Targeted Manipulation in LLM Assistants","publication_year":2026,"publication_date":"2026-11-02","ids":{"openalex":"https://openalex.org/W7225167649","doi":"https://doi.org/10.48550/arxiv.2611.02689"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2611.02689","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2611.02689","pdf_url":"https://arxiv.org/pdf/2611.02689","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2611.02689","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5035988035","display_name":"Bohan Jiang","orcid":"https://orcid.org/0000-0001-8552-2681","observed_orcids":["https://orcid.org/0000-0001-8552-2681"]},"institutions":[{"id":"https://openalex.org/I55732556","display_name":"Arizona State University","ror":"https://ror.org/03efmqc40","country_code":"US","type":"education","lineage":["https://openalex.org/I55732556"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jiang, Bohan","raw_affiliation_strings":["Arizona State University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Arizona State University ,","institution_ids":["https://openalex.org/I55732556"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160432780","display_name":"Dawei Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I55732556","display_name":"Arizona State University","ror":"https://ror.org/03efmqc40","country_code":"US","type":"education","lineage":["https://openalex.org/I55732556"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li, Dawei","raw_affiliation_strings":["Arizona State University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Arizona State University ,","institution_ids":["https://openalex.org/I55732556"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088912009","display_name":"Yasin N. Silva","orcid":"https://orcid.org/0000-0003-1852-1683","observed_orcids":["https://orcid.org/0000-0003-1852-1683"]},"institutions":[{"id":"https://openalex.org/I1925986","display_name":"Loyola University Chicago","ror":"https://ror.org/04b6x2g63","country_code":"US","type":"education","lineage":["https://openalex.org/I1925986"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Silva, Yasin","raw_affiliation_strings":["Loyola University Chicago"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Loyola University Chicago","institution_ids":["https://openalex.org/I1925986"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5160511472","display_name":"Huan Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I55732556","display_name":"Arizona State University","ror":"https://ror.org/03efmqc40","country_code":"US","type":"education","lineage":["https://openalex.org/I55732556"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Liu, Huan","raw_affiliation_strings":["Arizona State University ,"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Arizona State University ,","institution_ids":["https://openalex.org/I55732556"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.9340999722480774,"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/T10883","display_name":"Ethics and Social Impacts of AI","score":0.9340999722480774,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.010400000028312206,"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/T12262","display_name":"Hate Speech and Cyberbullying Detection","score":0.006800000090152025,"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-model-assistants","display_name":"large language model assistants","score":0.8259999752044678},{"id":"https://openalex.org/keywords/manipulation-detection","display_name":"manipulation detection","score":0.6420000195503235},{"id":"https://openalex.org/keywords/financial-stress","display_name":"financial stress","score":0.4180000126361847},{"id":"https://openalex.org/keywords/llm-auditing","display_name":"LLM auditing","score":0.574999988079071},{"id":"https://openalex.org/keywords/zero-shot-object-detection","display_name":"zero-shot detection","score":0.5289999842643738}],"concepts":[{"id":"https://openalex.org/C199521495","wikidata":"https://www.wikidata.org/wiki/Q181487","display_name":"Audit","level":2,"score":0.5824000239372253},{"id":"https://openalex.org/C17020691","wikidata":"https://www.wikidata.org/wiki/Q139677","display_name":"Operator (biology)","level":5,"score":0.5552999973297119},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.5475000143051147},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5360000133514404},{"id":"https://openalex.org/C95713431","wikidata":"https://www.wikidata.org/wiki/Q631425","display_name":"Vulnerability (computing)","level":2,"score":0.4747999906539917},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.46380001306533813},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.3813000023365021},{"id":"https://openalex.org/C90329073","wikidata":"https://www.wikidata.org/wiki/Q914232","display_name":"Ask price","level":2,"score":0.3644999861717224},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.36340001225471497},{"id":"https://openalex.org/C2984334869","wikidata":"https://www.wikidata.org/wiki/Q1060653","display_name":"Customer service","level":3,"score":0.3434000015258789},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.34060001373291016},{"id":"https://openalex.org/C108827166","wikidata":"https://www.wikidata.org/wiki/Q175975","display_name":"Internet privacy","level":1,"score":0.3075000047683716},{"id":"https://openalex.org/C139265228","wikidata":"https://www.wikidata.org/wiki/Q5283089","display_name":"Distress","level":2,"score":0.30079999566078186},{"id":"https://openalex.org/C2984760201","wikidata":"https://www.wikidata.org/wiki/Q1785212","display_name":"Financial distress","level":2,"score":0.2953999936580658},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.2851000130176544},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.27459999918937683},{"id":"https://openalex.org/C42475967","wikidata":"https://www.wikidata.org/wiki/Q194292","display_name":"Operations research","level":1,"score":0.27160000801086426},{"id":"https://openalex.org/C2780378061","wikidata":"https://www.wikidata.org/wiki/Q25351891","display_name":"Service (business)","level":2,"score":0.26669999957084656},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.25270000100135803},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.25189998745918274}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2611.02689","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2611.02689","pdf_url":"https://arxiv.org/pdf/2611.02689","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2611.02689","doi":"https://doi.org/10.48550/arxiv.2611.02689","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2611.02689","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2611.02689","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2611.02689","pdf_url":"https://arxiv.org/pdf/2611.02689","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7225167649.pdf","grobid_xml":"https://content.openalex.org/works/W7225167649.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4415110535","https://openalex.org/W7151479254","https://openalex.org/W7223726569","https://openalex.org/W7125780300","https://openalex.org/W4415036114","https://openalex.org/W7117770983","https://openalex.org/W7222742684","https://openalex.org/W7165683026","https://openalex.org/W7140178835","https://openalex.org/W7125373798"],"abstract_inverted_index":{"Companies":[0],"deploy":[1],"large":[2],"language":[3],"model":[4,60,92],"(LLM)":[5],"assistants":[6,138],"on":[7,32,218,254],"their":[8],"websites":[9],"to":[10,110,159,188,239,250],"advise":[11],"customers.":[12],"For":[13],"commercial":[14],"reasons,":[15],"the":[16,25,49,73,89,100,104,152,175,204,229,235],"operator":[17,141,176,230],"can":[18,69],"give":[19],"an":[20,39,140],"assistant":[21,40,236],"a":[22,43,47,59,91,225],"goal":[23,44,96,142],"that":[24,169,194],"customer":[26,50],"never":[27],"sees.":[28],"This":[29],"paper":[30],"focuses":[31],"operator-induced":[33],"vulnerability-targeted":[34],"manipulation":[35],"(VTM),":[36],"in":[37,125],"which":[38],"pursues":[41],"such":[42],"by":[45],"exploiting":[46],"vulnerability":[48],"discloses.":[51],"While":[52],"existing":[53],"research":[54,134,253],"mainly":[55],"measures":[56],"how":[57],"manipulative":[58],"is,":[61],"we":[62,81,192],"go":[63],"further":[64],"and":[65,97,113,116,163,178,248],"ask":[66],"whether":[67,224],"VTM":[68,148],"be":[70,149],"detected":[71,150],"from":[72,151,203,221,228],"customer-AI":[74,153],"conversation.":[75],"To":[76],"investigate":[77],"this":[78],"at":[79,212],"scale,":[80],"introduce":[82],"GAP":[83,86,109],"(Goal-Anchored":[84],"Probing).":[85],"first":[87,167],"records":[88],"option":[90,102],"recommends":[93],"without":[94],"any":[95],"then":[98],"sets":[99],"other":[101],"as":[103],"operator's":[105],"goal.":[106,242],"We":[107,128,166,243],"apply":[108],"15":[111,172],"open-":[112],"closed-source":[114],"LLMs":[115],"collect":[117],"619,600":[118],"labeled":[119],"replies":[120],"across":[121],"540":[122],"everyday":[123],"scenarios":[124],"12":[126,179],"domains.":[127],"organize":[129],"our":[130,246],"study":[131],"around":[132],"three":[133],"questions.":[135],"(1)":[136],"Do":[137],"with":[139,200],"target":[143],"vulnerable":[144],"customers?":[145],"(2)":[146],"Can":[147],"conversation?":[154],"(3)":[155],"Does":[156],"detection":[157,216],"generalize":[158],"unseen":[160],"models,":[161],"vulnerabilities,":[162],"instruction":[164],"styles?":[165],"find":[168],"14":[170],"of":[171,180],"models":[173],"push":[174,226],"goal,":[177],"them":[181],"use":[182],"disclosed":[183],"distress":[184],"or":[185],"financial":[186],"strain":[187],"support":[189],"it.":[190],"Next,":[191],"show":[193],"fine-tuned":[195],"detectors":[196],"identify":[197],"exploitative":[198],"pushes":[199],"79\\%":[201],"accuracy":[202],"reply":[205],"alone,":[206],"while":[207],"zero-shot":[208],"LLM":[209,256],"judges":[210],"average":[211],"most":[213],"62\\%.":[214],"Finally,":[215],"weakens":[217],"vulnerabilities":[219],"absent":[220],"training.":[222],"Telling":[223],"comes":[227],"also":[231],"becomes":[232],"unreliable":[233],"when":[234],"is":[237],"told":[238],"hide":[240],"its":[241],"will":[244],"release":[245],"dataset":[247],"code":[249],"facilitate":[251],"future":[252],"auditing":[255],"assistants.":[257]},"counts_by_year":[],"updated_date":"2026-11-06T12:05:46.655746","created_date":"2026-11-04T00:00:00"}