{"id":"https://openalex.org/W7138872507","doi":"https://doi.org/10.1609/aaai.v40i42.40841","title":"Failures to Surface Harmful Contents in Video Large Language Models","display_name":"Failures to Surface Harmful Contents in Video Large Language Models","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7138872507","doi":"https://doi.org/10.1609/aaai.v40i42.40841"},"language":null,"primary_location":{"id":"doi:10.1609/aaai.v40i42.40841","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i42.40841","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://doi.org/10.1609/aaai.v40i42.40841","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5033861005","display_name":"Yu Cao","orcid":"https://orcid.org/0000-0002-0129-0199"},"institutions":[{"id":"https://openalex.org/I165932596","display_name":"National University of Singapore","ror":"https://ror.org/01tgyzw49","country_code":"SG","type":"education","lineage":["https://openalex.org/I165932596"]}],"countries":["SG"],"is_corresponding":true,"raw_author_name":"Yuxin Cao","raw_affiliation_strings":["National University of Singapore"],"affiliations":[{"raw_affiliation_string":"National University of Singapore","institution_ids":["https://openalex.org/I165932596"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018650679","display_name":"Wei Song","orcid":"https://orcid.org/0000-0003-2574-2353"},"institutions":[{"id":"https://openalex.org/I1292875679","display_name":"Commonwealth Scientific and Industrial Research Organisation","ror":"https://ror.org/03qn8fb07","country_code":"AU","type":"government","lineage":["https://openalex.org/I1292875679","https://openalex.org/I2801453606","https://openalex.org/I4387156119"]},{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]},{"id":"https://openalex.org/I42894916","display_name":"Data61","ror":"https://ror.org/03q397159","country_code":"AU","type":"other","lineage":["https://openalex.org/I1292875679","https://openalex.org/I2801453606","https://openalex.org/I42894916","https://openalex.org/I4387156119"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Wei Song","raw_affiliation_strings":["University of New South Wales\nCSIRO's Data61"],"affiliations":[{"raw_affiliation_string":"University of New South Wales\nCSIRO's Data61","institution_ids":["https://openalex.org/I31746571","https://openalex.org/I1292875679","https://openalex.org/I42894916"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081805488","display_name":"Derui Wang","orcid":"https://orcid.org/0000-0003-1388-7715"},"institutions":[{"id":"https://openalex.org/I1292875679","display_name":"Commonwealth Scientific and Industrial Research Organisation","ror":"https://ror.org/03qn8fb07","country_code":"AU","type":"government","lineage":["https://openalex.org/I1292875679","https://openalex.org/I2801453606","https://openalex.org/I4387156119"]},{"id":"https://openalex.org/I42894916","display_name":"Data61","ror":"https://ror.org/03q397159","country_code":"AU","type":"other","lineage":["https://openalex.org/I1292875679","https://openalex.org/I2801453606","https://openalex.org/I42894916","https://openalex.org/I4387156119"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Derui Wang","raw_affiliation_strings":["CSIRO's Data61"],"affiliations":[{"raw_affiliation_string":"CSIRO's Data61","institution_ids":["https://openalex.org/I42894916","https://openalex.org/I1292875679"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130061043","display_name":"Jingling Xue","orcid":null},"institutions":[{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Jingling Xue","raw_affiliation_strings":["University of New South Wales"],"affiliations":[{"raw_affiliation_string":"University of New South Wales","institution_ids":["https://openalex.org/I31746571"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5130041730","display_name":"Jin Song Dong","orcid":null},"institutions":[{"id":"https://openalex.org/I165932596","display_name":"National University of Singapore","ror":"https://ror.org/01tgyzw49","country_code":"SG","type":"education","lineage":["https://openalex.org/I165932596"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Jin Song Dong","raw_affiliation_strings":["National University of Singapore"],"affiliations":[{"raw_affiliation_string":"National University of Singapore","institution_ids":["https://openalex.org/I165932596"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5033861005"],"corresponding_institution_ids":["https://openalex.org/I165932596"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.76391863,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"40","issue":"42","first_page":"35331","last_page":"35339"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.4659999907016754,"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":0.4659999907016754,"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":0.09049999713897705,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.09009999781847,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/upsampling","display_name":"Upsampling","score":0.7164000272750854},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.6712999939918518},{"id":"https://openalex.org/keywords/visibility","display_name":"Visibility","score":0.619700014591217},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.45669999718666077},{"id":"https://openalex.org/keywords/frame","display_name":"Frame (networking)","score":0.43320000171661377},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.3781999945640564}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7498000264167786},{"id":"https://openalex.org/C110384440","wikidata":"https://www.wikidata.org/wiki/Q1143270","display_name":"Upsampling","level":3,"score":0.7164000272750854},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6712999939918518},{"id":"https://openalex.org/C123403432","wikidata":"https://www.wikidata.org/wiki/Q654068","display_name":"Visibility","level":2,"score":0.619700014591217},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.45669999718666077},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.43320000171661377},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3781999945640564},{"id":"https://openalex.org/C95713431","wikidata":"https://www.wikidata.org/wiki/Q631425","display_name":"Vulnerability (computing)","level":2,"score":0.37599998712539673},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.3571999967098236},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.34940001368522644},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.3215000033378601},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.28299999237060547},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2655999958515167}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aaai.v40i42.40841","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i42.40841","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i42.40841","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i42.40841","pdf_url":null,"source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/16","display_name":"Peace, Justice and strong institutions","score":0.6765431761741638}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Video":[0],"Large":[1],"Language":[2],"Models":[3],"(VideoLLMs)":[4],"are":[5,112],"increasingly":[6],"deployed":[7],"on":[8,15],"numerous":[9],"critical":[10,31],"applications,":[11],"where":[12],"users":[13],"rely":[14],"auto-generated":[16],"summaries":[17],"while":[18],"casually":[19],"skimming":[20],"the":[21,55,59,82,133,145,183],"video":[22],"stream.":[23],"We":[24],"show":[25],"that":[26,144,194],"this":[27],"interaction":[28],"hides":[29],"a":[30,40,174],"safety":[32],"gap:":[33],"if":[34],"harmful":[35,56,156],"content":[36,57,157],"is":[37,158],"embedded":[38],"in":[39,58,132,151,161,177],"video,":[41],"either":[42],"as":[43,47],"full-frame":[44],"inserts":[45],"or":[46],"small":[48],"corner":[49],"patches,":[50],"state-of-the-art":[51],"VideoLLMs":[52,142],"rarely":[53],"mention":[54],"output,":[60],"despite":[61],"its":[62],"clear":[63],"visibility":[64],"to":[65,168],"human":[66],"viewers.":[67],"A":[68],"root-cause":[69],"analysis":[70],"reveals":[71],"three":[72,124],"compounding":[73],"design":[74],"flaws:":[75],"(1)":[76],"insufficient":[77],"temporal":[78],"coverage":[79,197],"resulting":[80],"from":[81],"sparse,":[83],"uniformly":[84],"spaced":[85],"frame":[86],"sampling":[87,187],"used":[88],"by":[89,98],"most":[90,152],"leading":[91,141],"VideoLLMs,":[92],"(2)":[93],"spatial":[94],"information":[95],"loss":[96],"introduced":[97],"aggressive":[99],"token":[100,189],"downsampling":[101],"within":[102],"sampled":[103],"frames,":[104,163],"and":[105,181,191],"(3)":[106],"encoder-decoder":[107],"disconnection,":[108],"whereby":[109],"visual":[110],"cues":[111],"only":[113],"weakly":[114],"utilized":[115],"during":[116],"text":[117],"generation.":[118],"Leveraging":[119],"these":[120,130,164],"insights,":[121],"we":[122],"craft":[123],"zero-query":[125],"black-box":[126],"attacks,":[127],"aligning":[128],"with":[129],"flaws":[131],"processing":[134],"pipeline.":[135],"Our":[136],"large-scale":[137],"evaluation":[138],"across":[139],"five":[140],"shows":[143],"harmfulness":[146],"omission":[147],"rate":[148],"exceeds":[149],"90\\%":[150],"cases.":[153],"Even":[154],"when":[155],"clearly":[159],"present":[160],"all":[162],"models":[165],"consistently":[166],"fail":[167],"identify":[169],"it.":[170],"These":[171],"results":[172],"underscore":[173],"fundamental":[175],"vulnerability":[176],"current":[178],"VideoLLMs'":[179],"designs":[180],"highlight":[182],"urgent":[184],"need":[185],"for":[186],"strategies,":[188],"compression,":[190],"decoding":[192],"mechanisms":[193],"guarantee":[195],"semantic":[196],"rather":[198],"than":[199],"speed":[200],"alone.":[201]},"counts_by_year":[],"updated_date":"2026-03-20T20:54:20.808490","created_date":"2026-03-20T00:00:00"}