{"id":"W4386982649","doi":"10.1007/s10664-023-10380-1","title":"Is GitHub’s Copilot as bad as humans at introducing vulnerabilities in code?","year":2023,"lang":"en","type":"article","venue":"Empirical Software Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":101,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Vulnerability (computing); Code (set theory); Process (computing); Computer security; Perspective (graphical); Secure coding; Software; Software engineering; Software security assurance; Artificial intelligence; Operating system; Information security; Programming language","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","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0009188286,0.0004059611,0.0004711092,0.0007471671,0.0001750933,0.0002190436,0.001279467,0.0001956342,0.0002225075],"category_scores_gemma":[0.008167374,0.000430545,0.0001448494,0.002466303,0.00005658425,0.0005053682,0.001192301,0.00076359,0.001956843],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007065113,"about_ca_system_score_gemma":0.0001548969,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001605245,"about_ca_topic_score_gemma":0.00002222623,"domain_scores_codex":[0.996208,0.00006611568,0.0005365604,0.001044475,0.0009449642,0.001199955],"domain_scores_gemma":[0.9952915,0.00308552,0.00004562262,0.001117683,0.0001024623,0.0003571924],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00002797455,0.000132539,0.702413,0.0004167305,0.00009869167,0.0006664553,0.01066577,0.2427232,0.0009366137,0.001139554,0.03754087,0.003238657],"study_design_scores_gemma":[0.001648072,0.0006143534,0.7704953,0.0004618157,0.00001341314,0.000128142,0.0001028208,0.1269248,0.009269198,0.001748921,0.08645829,0.002134904],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9011679,0.0002737354,0.09115834,0.002091942,0.0009352678,0.0003150715,0.00001017983,0.003976048,0.00007155703],"genre_scores_gemma":[0.9809093,0.00004022772,0.01469587,0.0005557049,0.0003761377,0.0001934432,0.00001850408,0.0001296648,0.00308117],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1157984,"threshold_uncertainty_score":0.9998146,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03474644801521013,"score_gpt":0.3154596742478399,"score_spread":0.2807132262326297,"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."}}