{"id":"W4415282304","doi":"10.3390/info16100904","title":"Can We Trust AI Content Detection Tools for Critical Decision-Making?","year":2025,"lang":"en","type":"article","venue":"Information","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia, Okanagan Campus; Government of Manitoba","funders":"","keywords":"Government (linguistics); Robustness (evolution); Precision and recall; Recall; Reliability (semiconductor); Content analysis","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":[],"consensus_categories":[],"category_scores_codex":[0.0003168017,0.00009124717,0.0001016878,0.0002269465,0.0002538921,0.000757469,0.0003815787,0.0000771598,0.000008336473],"category_scores_gemma":[0.002406414,0.0000884745,0.00006166479,0.0003837971,0.00003268217,0.00446609,0.00009130458,0.00009309503,0.00009030242],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001575515,"about_ca_system_score_gemma":0.00009743066,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003342444,"about_ca_topic_score_gemma":0.00007499541,"domain_scores_codex":[0.9990203,0.00001813978,0.000419779,0.0001272237,0.0001929571,0.0002215434],"domain_scores_gemma":[0.9984218,0.000740489,0.00006616861,0.0002951411,0.0004349838,0.00004145623],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002119031,0.00001002447,0.00001345073,0.00002376966,0.00000349004,3.424171e-7,0.0004938686,0.0003312871,0.000125794,0.3090114,0.000479988,0.6894854],"study_design_scores_gemma":[0.0001942683,0.0001445392,0.0004515331,0.0001801508,0.000009139117,0.000007147952,0.001002062,0.6247033,0.04519354,0.2868497,0.04106124,0.0002033617],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004543378,0.00001856268,0.9865245,0.005451624,0.001066026,0.0003656096,0.000005940688,0.0001417755,0.001882628],"genre_scores_gemma":[0.9766305,0.000004406831,0.02052267,0.002665821,0.00003395624,0.0001004568,0.000003136794,0.000002455283,0.00003655862],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9720871,"threshold_uncertainty_score":0.7304291,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03774558943092063,"score_gpt":0.3273950318151122,"score_spread":0.2896494423841915,"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."}}