AI law and the newsroom, October 31–November 3: four newspapers sue Brave over the index it resells, and Europe's court adviser takes up chatbot summaries
Two court developments for newsrooms: four US newspapers sue Brave Software over the search index it resells to AI companies, and the EU Court of Justice's adviser is due to give the first view on whether a chatbot's summary of a news article needs the publisher's permission.
This issue covers 31 October to 3 November 2026, the days since the last one, in two items. In New York, four newspapers sued a browser company over the search index it builds and resells. In Luxembourg, an adviser to the European Union's top court was due to state a view on whether a chatbot's summary of a news article needs the publisher's permission.
Two questions run through both. What can a newsroom do about a crawler that never gives its name? And what would a European ruling on summaries change for publishers who sell subscriptions?
United States: four newspapers sue the index underneath the answers
The New York Times Company sued Brave Software on Monday 2 November 2026, in the federal court for the Southern District of New York. 1 Hours later the publishers of the New York Daily News, the Chicago Tribune and the Orlando Sentinel filed in the same court. 2
The Times case is No. 1:26-cv-9549. The second is No. 1:26-cv-9554.
Both complaints aim at the same piece of infrastructure. They describe the Brave Index, in Brave's own words, as "over 30 billion webpages". 1 The alleged chain has four steps: Brave copies articles into that index, its AI answers reproduce them, it sells access to the index, and its browser offers readers a way past paywalls. 12
The crawler gives no name. Brave has "never disclosed the name of its 'bot'", the Times says, so a publisher cannot write a robots.txt rule that names it. 1 Brave's own help page confirms the practice without apology. Its crawler "does not advertise a differentiated user agent because we must avoid discrimination from websites that allow only Google to crawl them". 3
robots.txt will not keep it out. The same page tells site owners that "robots.txt is not used to prevent a page from being indexed", and sends them to the noindex directive instead, which takes effect once Brave re-fetches the page. 3 The only route Brave documents is a directive that also removes the page from search results.
Brave offers one indirect path: if Googlebot cannot crawl a page, Brave's crawler will not crawl it either. 3 For a site that lives on search traffic, that is no choice at all.
The index is sold on. The complaints name five customers of Brave's Search API: Anthropic, Perplexity, Mistral, xAI and Cohere. 12 Brave's marketing, quoted in the Times complaint, calls the API "trusted by the world's top AI labs". 1
The paywall setting is a separate claim. Brave's browser carries a "Paywall blocker" setting that the complaints say lets a reader past a publisher's paywall without subscribing, and that Brave promotes as a way to "Bypass Paywalls". 1
That allegation carries the two Digital Millennium Copyright Act counts, and it does not depend on the fair-use argument. Those counts turn on whether the paywall is a technological measure controlling access, and whether the feature exists mainly to get round it. 1
The three newspapers add an advertising claim. They say Brave's browser blocks their ads by default and sells its own in their place. 2
Both complaints bring five counts: direct, contributory and vicarious infringement of copyright under 17 U.S.C. §501, circumvention and trafficking under 17 U.S.C. §1201(a)(1)(A) and (a)(2), and unjust enrichment under New York law. 12
The scale claimed differs. The Times puts the works covered by its registration exhibit at over 10 million. 1 The three newspapers count more than 2 million articles in registered back issues. 2 Both complaints cite Brave's 2025 revenue of more than $100 million and close to 130 million monthly users. 12
Amount sought: not stated. Both ask for statutory and actual damages, restitution of profits, a permanent injunction, costs and fees, and both demand a jury. 12
Nothing here binds a newsroom yet. A complaint is a set of allegations, Brave has not answered, and no court has ruled. Brave did not respond to requests for comment on the day. 4
Graham James, a Times spokesperson, said the company had asked Brave repeatedly to stop and filed "to hold Brave accountable for stealing our content to power its own products and to provide it to some of the most prominent and well resourced generative AI companies in the world". 4 Steven Lieberman, the lawyer for the three newspapers, framed it as fencing: "Buying stolen goods secondhand doesn't make them legitimate." 4
Brave has been on the other side of this question before. In March 2025 it sued News Corporation in the Northern District of California, asking a court to declare its web indexing fair use. 5 News Corp countersued in July 2026, making the same two allegations these newspapers now make: that Brave hides its crawlers so publishers cannot block them, and that it sells near-verbatim summaries of their articles to AI companies. 6
Three things follow for a newsroom. The defendant here is the supplier of the index rather than the company that trains the model, and the complaints name the five buyers without suing them. 1 The DMCA counts offer a route that does not require winning the fair-use argument first. 1 And the unnamed crawler explains why robots.txt rules have looked ineffective against this particular visitor. 3
European Union: the court adviser takes up the chatbot summary
On 3 November 2026 an Advocate General of the Court of Justice of the European Union was scheduled to deliver an opinion in Like Company v Google Ireland Limited, case C-250/25. 7 The opinion had not been published when this issue was prepared, and the Court's case file still listed the case as pending. 8
The case began in Hungary. Like Company runs news portals, among them balatonkornyeke.hu, and claims the rights of a press publisher. It sued Google Ireland over answers that Gemini produced about one of its articles, on a Hungarian singer and a plan to keep dolphins beside Lake Balaton. 8
The Budapest Környéki Törvényszék referred four questions, and lodged the reference on 3 April 2025. 9
The first asks whether displaying, in a chatbot's answer, a text partly identical to a press publisher's page is a communication to the public, where the passage is long enough to be protected by Article 15 of Directive (EU) 2019/790. 9 The second asks whether training the model is a reproduction of the works used. 9
The third, which follows from the second, asks whether the text and data mining exception in Article 4 of the same directive covers that reproduction. 9 The fourth asks whether an answer that follows a user's instruction matching an article is a reproduction by the chatbot's provider. 9
Two provisions set the range of answers. Article 15 gives press publishers a two-year right over the online use of their press publications by information society service providers, with hyperlinks and "individual words or very short extracts" outside it. 10 Article 4 lets commercial text and data mining run unless rightholders have reserved their rights "in an appropriate manner". 10
What each answer would open. If a summary inside an answer is a communication to the public, a publisher can license or sue over the output as well as over the training. If training is reproduction that Article 4 does not cover where rights are reserved, the reservation becomes the operative fact for every model provider selling into the EU.
The second thread ties back to the last issue. Article 53(1)(c) of the AI Act already requires providers of general-purpose models to identify and comply with a reservation made under Article 4(3), and the General-Purpose AI Code of Practice names robots.txt as the place to express it. 11
Nothing here binds anyone yet. An Advocate General's opinion is advisory, the Court may depart from it, and no date is set for the judgment. 7 A judgment, once given, would interpret directives that already apply, and courts in all 27 member states would have to read them that way.
Also on 3 November, the Court hears argument in C-474/25, on whether German transparency rules for "media intermediary services" can be applied to a provider established in another member state, alongside Article 27 of the Digital Services Act. 7
The compliance clock
| Date | What happens | Who it binds | Source |
|---|---|---|---|
| 3 November 2026 | Advocate General's opinion in C-250/25, Like Company | Nobody; the Court need not follow it | Court of Justice case file |
| 10 November 2026 | European Commission's targeted consultation on copyright and technology closes, extended from 3 November | Anyone who chooses to respond | European Commission |
| 30 November 2026 | Australia's Joint Select Committee on Artificial Intelligence reports, with AI and copyright in its terms of reference | The committee reports; the government decides what follows | Parliament of Australia |
| 2 December 2026 | EU AI Act: Article 5(1) points (ba) and (bb), and Articles 5(1a) and (1b), begin to apply | Providers and deployers of the systems those provisions name | Article 113 |
| 2 December 2027 | EU AI Act: high-risk duties for systems classified under Article 6(2) and Annex III | Providers and deployers of those systems | Article 113 |
| 2 August 2028 | EU AI Act: high-risk duties for systems classified under Article 6(1) and Annex I | Providers and deployers of those systems | Article 113 |
The next hard dates
The nearest two dates need nothing from a newsroom except attention. The Commission's copyright consultation closes on 10 November, and Australia's AI committee reports on 30 November. 1213
The first date that does change an obligation is 2 December 2026, when the AI Act's restrictions on systems producing intimate material and child sexual abuse material begin to apply. 14
References
- 1
- 2
- 3Brave Search Crawler, Brave Search help
search.brave.com
- 4
- 5
- 6
- 7Geradin Partners' Monthly EU Litigation Briefing, November 2026
geradinpartners.com
- 8
- 9Request for a preliminary ruling from the Budapest Környéki Törvényszék, Case C-250/25
infocuria.curia.europa.eu
- 10
- 11Article 53: Obligations for providers of general-purpose AI models
ai-act-service-desk.ec.europa.eu
- 12Commission seeks feedback on challenges and way forward in the area of effect of technology on copyright
digital-strategy.ec.europa.eu
- 13
- 14Article 113: Entry into force and application
ai-act-service-desk.ec.europa.eu
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.