{"id":"W4411542027","doi":"10.1145/3715275.3732211","title":"Trustworthy ML Regulation as a Principal-Agent Problem","year":2025,"lang":"en","type":"article","venue":"","topic":"Auction Theory and Applications","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute; University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada; Government of Canada; Canadian Institute for Advanced Research","keywords":"Trustworthiness; Computer science; Principal (computer security); Computer security; Principal component analysis; Artificial intelligence","routes":{"ca_aff":true,"ca_fund":true,"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":["insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.001300337,0.00007100058,0.0001062476,0.000170884,0.0002463502,0.0001238016,0.0003316218,0.00004921859,0.006140062],"category_scores_gemma":[0.0003511912,0.00005028911,0.00006857568,0.001049541,0.00006073478,0.0001695802,0.00008685976,0.00006263769,0.003778272],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002672481,"about_ca_system_score_gemma":0.00007103191,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000009959418,"about_ca_topic_score_gemma":0.00001613274,"domain_scores_codex":[0.9986525,0.00009249101,0.0003994793,0.0003327543,0.0004055727,0.0001171592],"domain_scores_gemma":[0.998854,0.0002899596,0.0001059161,0.000526037,0.0001702116,0.00005388086],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00001268348,0.00003592758,0.0008778844,0.000001071285,0.000005828977,1.840884e-7,0.00008504489,0.0006158802,0.0002564116,0.9284695,0.0226598,0.04697976],"study_design_scores_gemma":[0.00008781462,0.000008006046,0.006064546,0.000003323177,0.00000417207,0.000001366432,0.0002027056,0.0009299793,0.00145737,0.5604672,0.4307293,0.00004415499],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.06051979,0.00002277884,0.2364683,0.006711596,0.0001820081,0.0003716332,0.000001861099,0.0001349778,0.695587],"genre_scores_gemma":[0.7000508,0.000001905462,0.002153261,0.0005710875,0.00002937383,0.00005502169,0.000002066527,0.000002419256,0.2971341],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.639531,"threshold_uncertainty_score":0.9969974,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0590448122556027,"score_gpt":0.4013054912616431,"score_spread":0.3422606790060403,"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."}}