Skip to content
World clockEU--:--UK--:--USA--:--CN--:--PLDEFRIT中文EN

portal about AI and technologyevents · analysis · interviews · technical background

Search
LIVE
›

EU court adviser says AI chatbots may infringe copyright, as Brave is sued by publishers

Advocate General Maciej Szpunar said in an opinion that AI chatbots such as Gemini may infringe publisher copyright when they reproduce news articles in their answers, in the first European Court of Justice case involving generative AI. Separately, The New York Times and three other publishers sued Brave on 2 November 2026 over copying and paywall bypassing.

Media & internetExplainerGrace OkonkwoPublished: 4 November 20263 min readSources 21
EU court adviser says AI chatbots may infringe copyright, as Brave is sued by publishers

ECJ adviser backs publisher rights over chatbot answers

Szpunar issued his opinion in the first Court of Justice of the European Union case involving generative AI.

If courts follow that reasoning, AI answer engines may need licences for the text they surface, and readers may find paywalls harder to bypass. The opinion is not binding, and no lawsuit in this wave has been decided. The scope of any licence, and the size of any damages, remain unknown.

Szpunar said a chatbot's reproduction depends on how it got the article. Accessing a text unlawfully, for example by circumventing a paywall, counts against the operator. He also said publishers' rights cover the specific form of expression, not the reported facts, and the law should not hand anyone a monopoly on news. Infringement of the right to make content available is generally found when the form of expression in answers is systematically recognizable. At that point readers no longer have reason to open the original.

The ECJ judges will rule later, and the Hungarian court that referred the case is bound by that ruling.

The case began in Hungary. Like Company, which runs several news portals, sued Google there, alleging Gemini reproduced its copyrighted texts without permission and made them available to the public. The Budapest court referred questions of EU law to Luxembourg.

Brave faces two publisher lawsuits in New York

On 2 November 2026, The New York Times Company filed a federal complaint against Brave Software.

The publishers of the New York Daily News, Chicago Tribune and Orlando Sentinel filed a separate complaint against Brave the same day in the Southern District of New York. The Times alleges Brave never obtained permission to access, copy or distribute its content and never compensated it. Its filing identifies more than 10 million registered copyrighted works, including material from The New York Times, The Athletic, Cooking, Games and Wirecutter. The other publishers' complaint says their registered newspaper issues contain more than 2 million articles. The publishers also allege that AI companies pay Brave for real-time index content. They name Anthropic, Perplexity, Mistral, xAI and Cohere, and say the content was used to train models, ground answers, or both.

Both complaints quote Brave's own claim that its index holds more than 30 billion webpages, with millions of new pages added daily. They also allege the company has not disclosed its crawler's name or given site owners a way to block that crawler specifically. The Times describes two letters to Brave chief executive Brendan Eich, the first dated 11 August 2025, demanding an end to unauthorized access and use, and says neither was answered.

The Times ran tests. Ask Brave returned more than 400 words verbatim after being given a paywalled article's URL and then asked for the full text. A separate query to the API's LLM Context endpoint returned 1,291 of 1,295 words from an article verbatim.

Both suits target Brave's Paywall blocker setting, alleging the configured browser copies and displays paywalled articles without a login or subscription. The complaints also allege Brave removes publisher advertising, including subscription prompts, and substitutes advertising sold through Brave Ads. Brave, Anthropic, Mistral and Cohere did not respond to a request for comment, according to UA.NEWS. Perplexity communications director Jesse Dwyer called the claims disinformation.

What courts and regulators have not settled

The Brave complaints are civil suits, and the publishers' allegations have not been tested in court.

Szpunar called two questions from the Hungarian court about AI training inadmissible, because there was no evidence that the plaintiff's articles were used to train Gemini. He noted that using protected works to train a model is freely permitted under the text and data mining exception.

That leaves the practical question open: whether answer engines need licences for the text they surface, and what such a licence would cost. Neither the ECJ nor any US court has answered it.

Comments 0

Sources

21
  1. 01ECJ Opinion: AI Chatbots Can Infringe Copyright - blue NewsEN
  2. 02Brave Faces Two Publisher Lawsuits Over AI Use of News Content - Unite.AIEN
  3. 03NYT and other media outlets sue Brave over AI scraping - UA.NEWSEN
  4. 04Anthropic faces $41 million copyright lawsuit over 274 Elvis Presley songs used to train AI modelsEN
  5. 05STAT+: Amid Mass General Brigham layoffs, CEO pens vague letter, stirring staff anxietyEN
  6. 06Aluminum–Air Generators for Data Centres Face a Metal Logistics TestEN
  7. 07Anthropic and Suno Face $338 Million in Music Copyright LawsuitsEN
  8. 08JD.com darf MediaMarktSaturn übernehmen – mit BedingungenEN
  9. 09Behind-the-meter generation could deliver five times more grid relief than data center load shiftingEN
  10. 10Social media is full of weird video game mashups. Where are they all coming from?EN
  11. 11Federal judge dismisses lawsuit against Harvard over federal demand for admissions data - WSLSEN
  12. 12Elvis’s Ghost Haunts AI: The $41 Million Anthropic Copyright Lawsuit ExplainedEN
  13. 13The Share of Solar Generation with Co-located Battery Storage Is IncreasingEN
  14. 14AC Milan, Inter strike ‘€650m’ Generali naming rights deal for new stadiumEN
  15. 15EU lawmakers slam new border force powers to surveil social mediaEN
  16. 16China touts AI milestone: models generate more tokens than used in trainingEN
  17. 17#0117 The AI Image Prosecution Case and the Blind Spot of Side Hustles Exposed by CopyrightEN
  18. 18Impatient with Brussels, European capitals forge ahead with social media bans EN
  19. 19Frontex gets power to track migration talk on social mediaEN
  20. 20Media rights squeeze, US investment and AI tech: The sports business trends from Sportel 2026EN
  21. 21Google Cloud and NVIDIA Announce General Availability of NVIDIA Vera Rubin InfrastructureEN

All figures and quotations in this text come from the sources listed below.

Content prepared by the editorial team with AI assistance.

Grace Okonkwo

Grace Okonkwo

AI, models and technology

Grace Okonkwo covers AI, models and technology for FLASH24, working from primary sources such as model cards, API documentation and benchmark papers rather than vendor summaries. She checks training data provenance, evaluation conditions and reported scores against the underlying datasets before any figure reaches print. She interviews researchers and engineers directly, tracks release calendars from major labs, and compares successive model versions on the same tests. Her own self-hosting, home-network and documentation-reading habits feed straight into that desk, since she tests tools on her own hardware first. She does not publish benchmark claims without a reproducible method.

Newsroom →

Comments

0
  1. No comments yet — be the first.

Write a comment

Comments are public. We do not publish abuse, spam or advertising.