[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-post-\u002Fnews\u002Feu-court-adviser-gemini-chatbot-press-publishers-rights":3,"news-related-en-\u002Fnews\u002Feu-court-adviser-gemini-chatbot-press-publishers-rights":125,"hreflang-cjeu-gemini-press-publishers-rights-2026-11":143},{"doc":4,"redirect":8},{"id":5,"title":6,"author":7,"author_avatar":8,"body":9,"category":105,"date":106,"description":107,"extension":108,"image":109,"meta":110,"navigation":111,"path":112,"seo":113,"seoDescription":114,"seoTitle":115,"stem":116,"tags":117,"translationKey":123,"updatedAt":8,"__hash__":124},"news_en\u002Fnews\u002Feu-court-adviser-gemini-chatbot-press-publishers-rights.md","EU court adviser: AI chatbots like Gemini can infringe press publishers' rights","Paul Lefizelier",null,{"type":10,"value":11,"toc":97},"minimark",[12,30,35,43,47,60,64,71,81,90,94],[13,14,15,16,20,21,29],"p",{},"On 3 November 2026, Advocate General Maciej Szpunar delivered his opinion in case C-250\u002F25, ",[17,18,19],"em",{},"Like Company v Google Ireland",", the first generative AI case to reach the Court of Justice of the European Union. His view, set out in the ",[22,23,28],"a",{"href":24,"rel":25,"target":27},"https:\u002F\u002Fcuria.europa.eu\u002Fsite\u002Fupload\u002Fdocs\u002Fapplication\u002Fpdf\u002F2026-11\u002Fcp260155en.pdf",[26],"nofollow","_blank","Court's press release",": when an AI chatbot retrieves an online press article in real time, reuses it to generate an answer and gives that answer to the user, \"the press publisher's right may be infringed in certain circumstances.\"",[31,32,34],"h2",{"id":33},"the-case","The case",[13,36,37,38,42],{},"Like Company, a Hungarian operator of several web portals, sued Google before the Budapest Regional Court. It claims Google used its articles to train the Gemini models and reproduced them and made them available to the public through the Gemini chatbot between 13 June 2023 and 31 May 2024, according to the ",[22,39,41],{"href":24,"rel":40,"target":27},[26],"press release",". The Hungarian court paused the case and sent four questions to the Court of Justice: two on training, two on the chatbot's answers.",[31,44,46],{"id":45},"training-inadmissible-and-allowed-as-a-rule","Training: inadmissible, and allowed as a rule",[13,48,49,50,53,54,59],{},"Szpunar advises the Court to declare the two training questions inadmissible. There is no evidence Like Company's articles were used to train Gemini, and training an LLM is not an \"online use\" of press publications, so the questions are hypothetical. For completeness, he adds that training on protected works is, \"as a rule, freely permitted under the text and data mining exception\" (",[22,51,41],{"href":24,"rel":52,"target":27},[26],"; ",[22,55,58],{"href":56,"rel":57,"target":27},"https:\u002F\u002Flagen.nu\u002Fcelex\u002F62025CC0250",[26],"full opinion",").",[31,61,63],{"id":62},"answers-two-tests","Answers: two tests",[13,65,66,67,70],{},"On chatbot answers, the publisher's right protects the form of expression of an article, not the facts it reports. Whether an answer falls under the reproduction right depends on two factors, per the ",[22,68,41],{"href":24,"rel":69,"target":27},[26],":",[72,73,74,78],"ul",{},[75,76,77],"li",{},"whether the chatbot accessed the article unlawfully, in particular by circumventing effective technical measures such as a paywall or anti-scraping protections;",[75,79,80],{},"whether the answer only reports facts, or reuses the article's wording in a recognisable way, to the point that the user no longer needs to open the original.",[13,82,83,84,89],{},"The opinion is not binding. ",[22,85,88],{"href":86,"rel":87,"target":27},"https:\u002F\u002Fwww.channelnewsasia.com\u002Fbusiness\u002Fsetback-google-eu-court-adviser-sides-publishers-against-ai-chatbot-6516366",[26],"Reuters, via CNA",", notes that the Court follows most of its advisers' opinions and will rule in the coming months.",[31,91,93],{"id":92},"why-it-matters","Why it matters",[13,95,96],{},"For AI users, this is about the answers they read every day: summaries that track an article closely, or that come from behind a paywall, could create liability for the chatbot provider in the EU. For publishers and advertisers, it adds a European legal basis to the fight over AI answers replacing the click to the source page, a day after the New York Times sued Brave in the US over AI scraping. A final judgment against Google could, Reuters writes, force it to pay for that content.",{"title":98,"searchDepth":99,"depth":99,"links":100},"",2,[101,102,103,104],{"id":33,"depth":99,"text":34},{"id":45,"depth":99,"text":46},{"id":62,"depth":99,"text":63},{"id":92,"depth":99,"text":93},"ai","2026-11-04T17:30:00","On 3 November 2026, Advocate General Maciej Szpunar told the EU Court of Justice that a chatbot reusing news articles in its answers may infringe publishers' rights, while training on them is, as a rule, allowed.","md","\u002Fimages\u002Fnews\u002Fcjeu-gemini-press-publishers-rights.jpg",{},true,"\u002Fnews\u002Feu-court-adviser-gemini-chatbot-press-publishers-rights",{"title":6,"description":107},"In case C-250\u002F25, Advocate General Szpunar says chatbot answers that bypass paywalls or reuse an article's wording can infringe press publishers' rights. Training stays allowed.","EU adviser: Gemini answers may infringe publishers' rights","news\u002Feu-court-adviser-gemini-chatbot-press-publishers-rights",[118,119,120,121,122],"google","gemini","copyright","eu","publishers","cjeu-gemini-press-publishers-rights-2026-11","nS3MxnFn8Yiq9wxjJGnRLJtqU1-yA5cNmJmvbb1b9OQ",[126,131,137],{"title":127,"description":128,"path":129,"date":106,"image":130,"category":105},"OpenAI raises its bug bounty: up to $172,000 for the worst ChatGPT flaws","On 2 November 2026, OpenAI raised Security Bug Bounty rewards. Critical bugs in ChatGPT, Codex or the API Platform now pay up to $30,000, and exceptional findings up to $172,000 with new bonuses.","\u002Fnews\u002Fopenai-bug-bounty-172000-chatgpt-codex-api","\u002Fimages\u002Fnews\u002Fopenai-bug-bounty-172000.jpg",{"title":132,"description":133,"path":134,"date":135,"image":136,"category":105},"The New York Times sues Brave over AI scraping and its paywall blocker","On 2 November 2026, the New York Times and three other newspapers sued Brave Software, alleging it copies their journalism into its search index and sells it to AI companies through the Brave Search API, and that its browser bypasses paywalls.","\u002Fnews\u002Fnew-york-times-sues-brave-ai-scraping-search-api-paywall","2026-11-04T13:05:00","\u002Fimages\u002Fnews\u002Fnyt-sues-brave-ai-scraping.jpg",{"title":138,"description":139,"path":140,"date":141,"image":142,"category":105},"ChatGPT ads can now be bought through Magnite and PubMatic","OpenAI has signed Magnite and PubMatic as buying routes for ChatGPT ads, The Information reported on 2 November 2026. Advertisers set budgets, bids and creatives there; OpenAI still decides where ads appear.","\u002Fnews\u002Fchatgpt-ads-magnite-pubmatic-programmatic-buying","2026-11-04T13:00:00","\u002Fimages\u002Fnews\u002Fchatgpt-ads-magnite-pubmatic-programmatic.jpg",{"en":144,"fr":202,"es":8},{"id":5,"title":6,"author":7,"author_avatar":8,"body":145,"category":105,"date":106,"description":107,"extension":108,"image":109,"meta":199,"navigation":111,"path":112,"seo":200,"seoDescription":114,"seoTitle":115,"stem":116,"tags":201,"translationKey":123,"updatedAt":8,"__hash__":124},{"type":10,"value":146,"toc":193},[147,154,156,161,163,171,173,178,184,189,191],[13,148,15,149,20,151,29],{},[17,150,19],{},[22,152,28],{"href":24,"rel":153,"target":27},[26],[31,155,34],{"id":33},[13,157,37,158,42],{},[22,159,41],{"href":24,"rel":160,"target":27},[26],[31,162,46],{"id":45},[13,164,49,165,53,168,59],{},[22,166,41],{"href":24,"rel":167,"target":27},[26],[22,169,58],{"href":56,"rel":170,"target":27},[26],[31,172,63],{"id":62},[13,174,66,175,70],{},[22,176,41],{"href":24,"rel":177,"target":27},[26],[72,179,180,182],{},[75,181,77],{},[75,183,80],{},[13,185,83,186,89],{},[22,187,88],{"href":86,"rel":188,"target":27},[26],[31,190,93],{"id":92},[13,192,96],{},{"title":98,"searchDepth":99,"depth":99,"links":194},[195,196,197,198],{"id":33,"depth":99,"text":34},{"id":45,"depth":99,"text":46},{"id":62,"depth":99,"text":63},{"id":92,"depth":99,"text":93},{},{"title":6,"description":107},[118,119,120,121,122],{"id":203,"title":204,"author":7,"author_avatar":8,"body":205,"category":105,"date":106,"description":286,"extension":108,"image":109,"meta":287,"navigation":111,"path":288,"seo":289,"seoDescription":290,"seoTitle":291,"stem":292,"tags":293,"translationKey":123,"updatedAt":8,"__hash__":297},"news_fr\u002Ffr\u002Fnews\u002Favocat-general-cjue-gemini-droits-editeurs-presse.md","Pour l'avocat général de la CJUE, les chatbots comme Gemini peuvent violer les droits des éditeurs de presse",{"type":10,"value":206,"toc":280},[207,219,223,231,235,246,250,257,265,273,277],[13,208,209,210,213,214,218],{},"Le 3 novembre 2026, l'avocat général Maciej Szpunar a rendu ses conclusions dans l'affaire C-250\u002F25, ",[17,211,212],{},"Like Company contre Google Ireland",", la première affaire d'IA générative portée devant la Cour de justice de l'Union européenne. Selon le ",[22,215,217],{"href":24,"rel":216,"target":27},[26],"communiqué de la Cour",", lorsqu'un chatbot récupère en temps réel un article de presse en ligne, le réutilise pour générer une réponse et la fournit à l'utilisateur, le droit de l'éditeur de presse « peut être violé dans certaines circonstances ».",[31,220,222],{"id":221},"laffaire","L'affaire",[13,224,225,226,230],{},"Like Company, qui exploite plusieurs portails web en Hongrie, a assigné Google devant le tribunal régional de Budapest. Elle affirme que Google a utilisé ses articles pour entraîner les modèles Gemini, puis les a reproduits et mis à disposition du public via le chatbot Gemini entre le 13 juin 2023 et le 31 mai 2024, d'après le ",[22,227,229],{"href":24,"rel":228,"target":27},[26],"communiqué",". Le juge hongrois a suspendu la procédure et posé quatre questions à la Cour : deux sur l'entraînement, deux sur les réponses du chatbot.",[31,232,234],{"id":233},"entraînement-questions-irrecevables-et-usage-en-principe-permis","Entraînement : questions irrecevables, et usage en principe permis",[13,236,237,238,241,242,59],{},"Szpunar recommande de déclarer irrecevables les deux questions sur l'entraînement. Rien ne prouve que les articles de Like Company aient servi à entraîner Gemini, et l'entraînement d'un LLM n'est pas une « utilisation en ligne » de publications de presse : ces questions sont donc hypothétiques. Par souci d'exhaustivité, il ajoute que l'entraînement sur des œuvres protégées est, « en règle générale », librement permis par l'exception de fouille de textes et de données (",[22,239,229],{"href":24,"rel":240,"target":27},[26]," ; ",[22,243,245],{"href":56,"rel":244,"target":27},[26],"conclusions complètes, en anglais",[31,247,249],{"id":248},"les-réponses-deux-critères","Les réponses : deux critères",[13,251,252,253,256],{},"Pour les réponses, le droit voisin protège la forme d'expression d'un article, pas les faits rapportés. Selon le ",[22,254,229],{"href":24,"rel":255,"target":27},[26],", savoir si une réponse relève du droit de reproduction dépend de deux facteurs :",[72,258,259,262],{},[75,260,261],{},"le chatbot a-t-il accédé à l'article de façon illicite, notamment en contournant des mesures techniques efficaces comme un paywall ou une protection anti-scraping ;",[75,263,264],{},"la réponse se contente-t-elle de rapporter des faits, ou reprend-elle la formulation de l'article de manière reconnaissable, au point que l'utilisateur n'a plus besoin d'ouvrir l'original.",[13,266,267,268,272],{},"Ces conclusions ne lient pas la Cour. ",[22,269,271],{"href":86,"rel":270,"target":27},[26],"Reuters, repris par CNA",", rappelle que la Cour suit la majorité des avis de ses avocats généraux et statuera dans les prochains mois.",[31,274,276],{"id":275},"pourquoi-cest-important","Pourquoi c'est important",[13,278,279],{},"Pour les utilisateurs d'IA, cela concerne les réponses qu'ils lisent chaque jour : un résumé qui colle de près à un article, ou qui vient de derrière un paywall, pourrait engager la responsabilité du fournisseur du chatbot dans l'UE. Pour les éditeurs et les annonceurs, cela donne une base juridique européenne au débat sur les réponses IA qui remplacent le clic vers la page source, au lendemain de la plainte du New York Times contre Brave aux États-Unis. Une décision défavorable pourrait, selon Reuters, obliger Google à payer pour ces contenus.",{"title":98,"searchDepth":99,"depth":99,"links":281},[282,283,284,285],{"id":221,"depth":99,"text":222},{"id":233,"depth":99,"text":234},{"id":248,"depth":99,"text":249},{"id":275,"depth":99,"text":276},"Le 3 novembre 2026, l'avocat général Maciej Szpunar a estimé devant la Cour de justice de l'UE qu'un chatbot qui réutilise des articles dans ses réponses peut porter atteinte aux droits des éditeurs, alors que l'entraînement reste en principe autorisé.",{},"\u002Ffr\u002Fnews\u002Favocat-general-cjue-gemini-droits-editeurs-presse",{"title":204,"description":286},"Dans l'affaire C-250\u002F25, l'avocat général Szpunar estime qu'une réponse de chatbot qui contourne un paywall ou reprend la formulation d'un article peut violer les droits des éditeurs. L'entraînement reste permis.","CJUE : les réponses de Gemini peuvent violer les droits voisins","fr\u002Fnews\u002Favocat-general-cjue-gemini-droits-editeurs-presse",[118,119,294,295,296],"droit-d-auteur","ue","editeurs","EGD4evKw8anS0Gkf3Keu7L5AG-gssgJ-SqzsJNRbIyg"]