{"id":"W4401415801","doi":"10.1109/access.2024.3440647","title":"A Systematic Literature Review on AI Safety: Identifying Trends, Challenges, and Future Directions","year":2024,"lang":"en","type":"article","venue":"IEEE Access","topic":"Adversarial Robustness in Machine Learning","field":"Computer Science","cited_by":30,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Chicoutimi","funders":"Mitacs","keywords":"Interpretability; Computer science; Software deployment; Notice; Autonomy; Risk analysis (engineering); Robustness (evolution); Adversarial system; Trustworthiness; Artificial intelligence; Data science; Computer security; Software engineering; Medicine","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01358069,0.001118584,0.003312122,0.01892103,0.0009137548,0.003426501,0.00183595,0.002254531,0.006721282],"category_scores_gemma":[0.08004707,0.0009685789,0.004184439,0.0163003,0.001549986,0.004933559,0.002227411,0.002418975,0.001000969],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002749964,"about_ca_system_score_gemma":0.01934349,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00514716,"about_ca_topic_score_gemma":0.01792273,"domain_scores_codex":[0.9906826,0.003264264,0.0031695,0.0007571882,0.001861267,0.0002651305],"domain_scores_gemma":[0.8740038,0.1031963,0.0100649,0.00174051,0.01006857,0.0009259299],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"systematic_review","study_design_scores_codex":[0.0001848213,0.00004882573,0.00204714,0.6047522,0.002002189,0.0001919109,0.0008065414,0.000465279,0.0002937092,0.004064286,0.02216182,0.3629813],"study_design_scores_gemma":[0.00004682607,0.0001251337,0.003290428,0.85074,0.007083021,0.0005132615,0.0008357303,0.0001985271,0.0001963524,0.00358891,0.1333337,0.00004798258],"study_design_candidate":"systematic_review","study_design_consensus":"systematic_review","genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.0005301076,0.9958406,0.0005020787,0.002098803,0.0001961619,0.00005375159,0.0002213468,0.00001436498,0.0005426309],"genre_scores_gemma":[0.004063104,0.9931726,0.0009965781,0.001191568,0.0001639906,0.00009946037,0.0002120504,0.00000879883,0.0000917923],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.01892103,"threshold_uncertainty_score":0.07182246,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02978114980185721,"score_gpt":0.339898914322889,"score_spread":0.3101177645210318,"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."}}