{"id":"W4304687331","doi":"10.20944/preprints202210.0157.v1","title":"Adversarial Artificial Intelligence in Insurance: From an Example to Some Potential Remedies","year":2022,"lang":"en","type":"preprint","venue":"Preprints.org","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Adversarial system; Underwriting; Intermediary; Actuarial science; Business; Robustness (evolution); Taxonomy (biology); Computer science; Computer security; Artificial intelligence; Finance","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","open_science","research_integrity","insufficient_payload"],"consensus_categories":["open_science"],"category_scores_codex":[0.002090166,0.0006761644,0.0008713359,0.0006459075,0.0003622357,0.0001921844,0.005867671,0.0004297935,0.0013224],"category_scores_gemma":[0.0008603889,0.0008024917,0.000253763,0.0007099921,0.0001358616,0.0009613856,0.01881355,0.003107681,0.0006094709],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005758674,"about_ca_system_score_gemma":0.0004925518,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01328765,"about_ca_topic_score_gemma":0.0004997594,"domain_scores_codex":[0.9925937,0.001015884,0.001217429,0.003057705,0.00127526,0.0008400435],"domain_scores_gemma":[0.9952947,0.0002733141,0.0005233524,0.003432421,0.0001411552,0.0003350494],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"observational","study_design_scores_codex":[0.0002854089,0.0003478147,0.08778136,0.00004661786,0.00008874365,0.0002540309,0.01390611,0.8613337,0.002785484,0.01489441,0.0000160763,0.01826027],"study_design_scores_gemma":[0.0006066949,0.0001694265,0.4707738,0.0002757121,0.00006019907,0.0000106869,0.001478592,0.2214233,0.007796894,0.292366,0.002397801,0.002640895],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6808556,0.0000522202,0.3090788,0.0009381502,0.007370048,0.0008127892,0.00003755454,0.0004614228,0.0003935009],"genre_scores_gemma":[0.9720879,0.0000380997,0.02538417,0.0003745006,0.001549803,0.0002924802,0.0001138381,0.00006790948,0.00009125722],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6399104,"threshold_uncertainty_score":0.9995905,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1381291682684165,"score_gpt":0.3642873726372042,"score_spread":0.2261582043687877,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}