{"article":{"annotation":{"article_id":"duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps","available_at":"2026-10-16T10:37:03.483222Z","geography_revision":"ef6ee0010dda5fe7b240fd514749b43e214c57da73222d2e40ec83b276dd0977","id":"c9a5cabdd0ca744d840e978e1907d3a745a0e8c9856f1a974cb4d9364bf3c64e","input_sha256":"be667451b8c6a0647b47bf8206ff671764ac24b94347a3597ed95001f8f2b9ea","limitations":["Rule-derived subjects are candidates, not verified editorial labels or calibrated model probabilities.","Coverage is distinct from event location. Unknown country/city/coordinates remain absent; no automatic geocoding.","Article geometries keep their stated precision and methodology; reproducing them is not independent coordinate verification."],"method":"controlled-rules-v1","places":[{"admin1":null,"city":null,"coordinate_note":null,"coordinate_reference_system":"OGC:CRS84","coordinate_source":null,"country":"US","evidence":[{"field":"country","language":"en","note":"Explicit preserved editorial country/region; no city or national centroid inferred","source_urls":["https://law.duke.edu/news/how-well-do-chatbots-and-humans-agree-legal-reasonableness"]}],"geometry":null,"id":"iso3166-1:US","kind":"country-or-area","labels":{"ar":"الولايات المتحدة الأمريكية","en":"United States of America","es":"Estados Unidos de América","fr":"États-Unis d’Amérique"},"match_status":"candidate","precision":"regional","role":"coverage","uncertainty_m":null}],"rules_sha256":"fb5ad3c27c91cb840f7a0a05f7f64a5cb26ad0374409ff9457f277c7092e22b4","schema_version":1,"subjects":[{"concept_id":"medtop:13000000","confidence":null,"confidence_method":null,"evidence":[{"field":"category+topics","language":"en","note":"Preserved editorial category/topic; broad candidate only","source_urls":["https://law.duke.edu/news/how-well-do-chatbots-and-humans-agree-legal-reasonableness"]}],"method":"controlled-rules-v1","rule_id":"technology-broad","status":"candidate"}],"taxonomy_revision":"1f2db31d811fca07e99b16d402c10e2cc2965c9df1c7b41d9be348b40a24e7fb","taxonomy_version":"IPTC Media Topics 2026-Q2"},"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=fr","category":"technology","content":{"key_points":[],"paragraphs":["Duke Law a résumé l’étude dans un article du 15 octobre. Les chercheurs ont comparé les réponses de 500 personnes et de 26 modèles d’IA — notamment des systèmes de ChatGPT, Claude, Gemini, Llama, Grok et DeepSeek — dans 25 scénarios juridiques demandant ce qui serait « raisonnable ». L’article avait été déposé comme prépublication sur arXiv et accepté à la conférence AAAI/ACM sur l’IA, l’éthique et la société.","Les chercheurs indiquent que les réponses des modèles correspondaient globalement aux jugements humains, mais se concentraient dans une fourchette plus étroite. Dans certains scénarios, les modèles accordaient davantage de latitude au gouvernement et aux entreprises qu’aux particuliers; leurs réponses ressemblaient aussi davantage à celles de répondants plus âgés, blancs, masculins et diplômés de l’enseignement supérieur. L’équipe estime que ces résultats doivent être vérifiés avec des participants plus diversifiés et davantage de chatbots."],"sections":[]},"corrections":[],"country":"US","data":{"blocks":[{"description":{"ar":"تقييمات لما يعد «معقولاً» عبر 25 سيناريو قانونياً.","de":"Bewertung dessen, was in 25 rechtlichen Szenarien „angemessen“ war.","en":"Assessments of what was “reasonable” across 25 legal scenarios.","es":"Evaluaciones de lo que era «razonable» en 25 escenarios legales.","fr":"Évaluations de ce qui était « raisonnable » dans 25 scénarios juridiques.","pt":"Avaliações do que seria «razoável» em 25 cenários jurídicos.","tr":"25 hukuki senaryoda “makul” görülen yanıtları değerlendirme."},"id":"study-record","items":[{"date":"2026-10-15","description":{"ar":"500 مشاركاً و26 نموذجاً للذكاء الاصطناعي و25 سيناريو.","de":"500 Teilnehmende, 26 KI-Modelle und 25 Szenarien.","en":"500 participants, 26 AI models and 25 scenarios.","es":"500 participantes, 26 modelos de IA y 25 escenarios.","fr":"500 participants, 26 modèles d’IA et 25 scénarios.","pt":"500 participantes, 26 modelos de IA e 25 cenários.","tr":"500 katılımcı, 26 AI modeli ve 25 senaryo."},"label":{"ar":"مقارنة إجابات البشر والنماذج","de":"Vergleich menschlicher Antworten und Modelle","en":"Human and model comparison","es":"Comparación entre respuestas humanas y modelos","fr":"Comparaison entre réponses humaines et modèles","pt":"Comparação de respostas humanas e modelos","tr":"İnsan ve model karşılaştırması"},"status":"occurred"}],"methodology":{"ar":"المنهج والنتائج وفق ملخص ديوك وبحث arXiv. وتقتصر النتائج على عينة الدراسة وسيناريوهاتها ولا تثبت أداءً موثقاً في أحكام المحاكم.","de":"Methoden und Ergebnisse laut Duke und arXiv-Arbeit. Die Ergebnisse sind auf Stichprobe und Szenarien begrenzt und belegen keine validierte Leistung in Gerichtsverfahren.","en":"Methods and findings as reported by Duke and the arXiv paper. Results are limited to the study sample and scenarios; they do not establish validated courtroom decision performance.","es":"Métodos y hallazgos según Duke y el artículo de arXiv. Los resultados se limitan a la muestra y escenarios del estudio; no validan el desempeño en tribunales.","fr":"Méthode et résultats selon Duke et l’article arXiv. Les résultats se limitent à l’échantillon et aux scénarios étudiés; ils ne valident pas une performance judiciaire en conditions réelles.","pt":"Métodos e resultados conforme Duke e o artigo do arXiv. Os resultados se limitam à amostra e aos cenários do estudo e não comprovam desempenho validado em tribunais.","tr":"Duke’un haber özeti ve arXiv makalesindeki yöntem ve bulgular. Sonuçlar çalışmadaki katılımcı örneklemi ve senaryolarla sınırlıdır; mahkemelerde doğrulanmış karar performansı göstermez."},"sourceUrls":["https://law.duke.edu/news/how-well-do-chatbots-and-humans-agree-legal-reasonableness"],"title":{"ar":"دراسة المعقولية القانونية","de":"Studie zu rechtlicher Angemessenheit","en":"Legal-reasonableness study","es":"Estudio de razonabilidad jurídica","fr":"Étude sur le caractère raisonnable en droit","pt":"Estudo de razoabilidade jurídica","tr":"Hukuki makullük çalışması"},"type":"events","updatedAt":"2026-10-16T10:20:41+00:00"}],"version":1},"data_downloads":[{"formats":{"csv":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps/data/study-record?lang=fr&format=csv","json":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps/data/study-record?lang=fr&format=json"},"id":"study-record"}],"data_types":["events"],"editorial_illustration":{"alt":"Illustration conceptuelle de la technologie, des processeurs et des informations connectées.","asset_id":"newsverum-v1-technology","avif_variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.avif","width":480},{"height":480,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-720.avif","width":720},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.avif","width":960},{"height":800,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1200.avif","width":1200},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.avif","width":1536}],"credit":"Newsverum","documentary":false,"height":1024,"kind":"symbolic-editorial-illustration","language":"fr","scope":"topic","url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.webp","width":480},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.webp","width":960},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","width":1536}],"width":1536},"event_id":"event-1800ea43-e9fe-550b-a7d6-b4e9ffeec2ba","event_period":"2026-10-15","event_period_basis":"event-or-reference-period","id":"duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps","image":null,"importance":0.6,"language":"fr","modified_at":"2026-10-16T10:36:24+00:00","publication_basis":"publication-record","published_at":"2026-10-16T10:36:24+00:00","reading_minutes":1,"region":"world","relations":[],"section_illustration":{"alt":"Illustration conceptuelle de la technologie, des processeurs et des informations connectées.","asset_id":"newsverum-v1-technology","avif_variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.avif","width":480},{"height":480,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-720.avif","width":720},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.avif","width":960},{"height":800,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1200.avif","width":1200},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.avif","width":1536}],"credit":"Newsverum","documentary":false,"height":1024,"kind":"symbolic-editorial-illustration","language":"fr","scope":"section","url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","variants":[{"height":320,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-480.webp","width":480},{"height":640,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-960.webp","width":960},{"height":1024,"url":"https://newsverum.com/assets/newsverum/illustrations/v1/technology-1536.webp","width":1536}],"width":1536},"slug":"duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps","sources":[{"checkedAt":"2026-10-16","name":"Duke University School of Law","publishedAt":"2026-10-15","url":"https://law.duke.edu/news/how-well-do-chatbots-and-humans-agree-legal-reasonableness"}],"summary":"Les chercheurs ont comparé les réponses de 500 personnes à celles de 26 modèles sur 25 scénarios juridiques; les modèles étaient plus uniformes et parfois plus indulgents envers les institutions.","text":"Les chercheurs ont comparé les réponses de 500 personnes à celles de 26 modèles sur 25 scénarios juridiques; les modèles étaient plus uniformes et parfois plus indulgents envers les institutions.\n\nDuke Law a résumé l’étude dans un article du 15 octobre. Les chercheurs ont comparé les réponses de 500 personnes et de 26 modèles d’IA — notamment des systèmes de ChatGPT, Claude, Gemini, Llama, Grok et DeepSeek — dans 25 scénarios juridiques demandant ce qui serait « raisonnable ». L’article avait été déposé comme prépublication sur arXiv et accepté à la conférence AAAI/ACM sur l’IA, l’éthique et la société.\n\nLes chercheurs indiquent que les réponses des modèles correspondaient globalement aux jugements humains, mais se concentraient dans une fourchette plus étroite. Dans certains scénarios, les modèles accordaient davantage de latitude au gouvernement et aux entreprises qu’aux particuliers; leurs réponses ressemblaient aussi davantage à celles de répondants plus âgés, blancs, masculins et diplômés de l’enseignement supérieur. L’équipe estime que ces résultats doivent être vérifiés avec des participants plus diversifiés et davantage de chatbots.","title":"Une étude de Duke relève des similitudes et écarts entre chatbots et jugements juridiques humains","topics":[],"translations":[{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=tr","language":"tr","url":"https://newsverum.com/tr/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=en","language":"en","url":"https://newsverum.com/en/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=ar","language":"ar","url":"https://newsverum.com/ar/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=fr","language":"fr","url":"https://newsverum.com/fr/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=es","language":"es","url":"https://newsverum.com/es/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=de","language":"de","url":"https://newsverum.com/de/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},{"api_url":"https://newsverum.com/api/v1/articles/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps?lang=pt","language":"pt","url":"https://newsverum.com/pt/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"}],"url":"https://newsverum.com/fr/news/duke-study-finds-chatbots-align-with-human-legal-reasoning-with-gaps"},"version":"1"}