On Monday, November 2, 2026, The New York Times sued Brave Software, the company behind the privacy-focused Brave browser and Brave Search, in federal court in Manhattan. The Times says Brave scraped its journalism, including paywalled articles, into a giant search index, then sold access to that index to AI companies. A few hours later, the publishers of the New York Daily News, the Chicago Tribune and the Orlando Sentinel filed a separate lawsuit making similar claims. In their complaint, Brave is "a burglar that misappropriates the content of publishers, and an AI 'fence' that sells stolen content."
The cases are The New York Times Co. v. Brave Software, Inc., No. 1:26-cv-09549, and Daily News, LP v. Brave Software, Inc., No. 1:26-cv-09554, both in the U.S. District Court for the Southern District of New York. The CourtListener docket for the Times case lists the cause as 17 U.S.C. § 501 copyright infringement, notes a plaintiff jury demand, and shows no judge assigned yet. Steven Lieberman of Rothwell, Figg, Ernst & Manbeck, the firm that represents the Daily News and other newspapers in their suits against OpenAI and Microsoft, appears as counsel in both cases. On filing day he also lodged copyright report forms with exhibits listing the works at issue.
Many AI copyright suits target model developers like OpenAI or Anthropic. These two target a company that sits one step earlier in the pipeline: the search-index supplier that, according to the publishers, delivers news content to AI companies.
What the publishers allege
According to the complaints, as described by Reuters, The Verge, TheWrap, The Next Web, MLex, IBTimes and PPC Land:
- A 30-billion-page index built by an unnamed crawler. Brave runs its own search index, the "Brave Index," with more than 30 billion webpages. The publishers say Brave has never published a name for its crawler. Brave's help pages say it avoids a separate user agent so that sites that allow only Googlebot will not block it. In practice, a publisher cannot block Brave in robots.txt without also blocking Google. The complaints present this as evidence of willfulness.
- Selling the index to AI companies. Brave sells index access through its Search API. The complaints describe endpoints for Web Search, News Search, "LLM Context" and Answers. Brave's own marketing says the API is often used for "training and grounding" large language models. The complaints name five API customers: Anthropic, Perplexity, Mistral, xAI and Cohere. They also cite a 2023 Brave blog post saying that Anthropic used the API in connection with Claude 2. None of the AI companies is a defendant. Lieberman said, "Buying stolen goods secondhand doesn't make them legitimate."
- Near-verbatim output. The complaints include test results. According to PPC Land's review, an LLM Context API call returned 1,291 of the 1,295 words of one Times article word for word. Brave's AI answer feature, Ask Brave (originally launched in 2023 as the Summarizer), allegedly reproduced more than 400 words of one Times article verbatim, more than 1,200 words of a Chicago Tribune article about data-center tax breaks, and nearly 600 words of a Daily News story about Central Park horse carriages.
- A "Paywall blocker." The DMCA claim targets a setting in Brave Shields called "Paywall blocker," which the complaints say is labeled with the phrase "Bypass Paywalls Clean Filters." They quote a Reddit user who said the feature let them read the Times "without interruption and without needing to set up an account."
- Ad substitution. Brave blocks publishers' ads by default and runs its own ad network, which shares revenue with users through Brave Rewards. The newspapers argue this lets Brave profit from their journalism while removing the ads that pay for it.
- Ignored warnings. The Times says it sent cease-and-desist letters to CEO Brendan Eich on August 11, 2025 and April 13, 2026. The second letter reportedly warned of enhanced statutory damages of up to $150,000 per work. A Times spokesperson told Reuters that Brave ignored both.
Each complaint brings five counts: direct, contributory and vicarious copyright infringement; circumvention of technological measures under DMCA § 1201(a)(1)(A) and anti-circumvention trafficking under § 1201(a)(2); and unjust enrichment under New York law. The publishers seek statutory and actual damages, disgorgement, a permanent injunction, § 1203 remedies including destruction of circumvention "devices," and a jury trial. Neither complaint states a dollar figure. The Times says it holds more than 10 million registered works, including Wirecutter and The Athletic. The three other newspapers claim more than 2 million registered articles.
These are allegations. Brave had not responded publicly or in court as of this writing. Brave has argued in the past that its indexing is lawful. In March 2025 it went to court first against News Corp, seeking a declaration that its indexing practices are fair use (Brave Software v. News Corporation, N.D. Cal.). News Corp has since asserted its own claims that Brave scrapes and resells journalism to AI developers.
The legal fault lines
1. Is a search index a library or a fence?
Brave's best precedents are the search-engine fair-use cases: Perfect 10 v. Amazon (9th Cir. 2007), Kelly v. Arriba Soft, and the Second Circuit's Authors Guild v. Google (2015). Those courts held that copying whole works to build a searchable index can be "highly transformative" when the output points users to the original instead of replacing it.
The publishers' framing aims at the line Google Books drew. Google showed only snippets and blocked users from reassembling full books. Here the complaints allege that the API and Ask Brave return substantially entire articles. The Second Circuit has since narrowed the "transformative" argument where the output substitutes for the original: Fox News v. TVEyes (2018) held that a media-monitoring service that let customers watch Fox clips was not fair use. After Andy Warhol Foundation v. Goldsmith (2023), courts also focus heavily on whether the secondary use serves the same commercial purpose as the original. Selling news content to AI companies that answer users' news questions looks a lot like a licensing market publishers already participate in.
2. Who is liable for what the AI companies do?
The contributory and vicarious counts treat Brave as a supplier whose customers commit the downstream infringement. To win, the publishers will need to show that Brave knew of and materially contributed to specific infringement, or that it profited from infringement it had the right and ability to control. The cease-and-desist letters matter here because they are evidence of knowledge. Brave will rely on Sony and Grokster-style arguments: a general-purpose search API has substantial non-infringing uses, and Brave cannot be liable for how each customer uses it.
3. Does a paywall blocker "circumvent" anything?
The § 1201 count may be the most important one for the wider internet. It requires that a "technological measure effectively controls access" to the work and that Brave or its users "circumvent" it. Many news paywalls are "soft." The full article is sent to the browser and then hidden by a script or overlay. Brave will argue that hiding an overlay isn't circumventing anything, because the publisher already delivered the content. The publishers will point to the statute's broad language and to the fact that the filter list is reportedly built expressly to defeat paywalls. The trafficking theory under § 1201(a)(2), which covers a product "primarily designed" for circumvention, could put ad-blocking and filter-list tools across the browser industry at risk. The complaints reportedly don't explain technically how the filter defeats the paywall, so expect Brave's motion to dismiss to focus on this point.
4. Is unjust enrichment preempted?
Section 301 of the Copyright Act preempts state claims that are "equivalent" to copyright. New York courts routinely dismiss unjust enrichment claims that are just copyright claims under another name. The publishers built this count around ad substitution, not copying, and that framing appears designed to survive preemption. Courts will need to decide whether blocking ads, by itself, gives rise to any state-law claim. Ad-blocking has historically survived legal challenges, especially in Germany's long Axel Springer v. Eyeo litigation.
Where this fits
Lieberman's firm also represents the Daily News and other papers in the consolidated newspaper litigation against OpenAI and Microsoft, and the Times has its own case there (NYT v. Microsoft). The Times also sued Perplexity last December (NYT v. Perplexity). By suing an index supplier, publishers are now targeting each layer of the retrieval-augmented generation pipeline: the model maker, the AI answer engine, and now the search index feeding both. A win against Brave would give publishers leverage over AI companies that never scraped a page themselves.
What this means for filmmakers and other creators
IP Feed readers who own film, TV, music or publishing rights should pay attention. The "index supplier" theory applies to any content that sits on the open web, not just news.
- Register your works. The Times can pursue statutory damages of up to $150,000 per work because it registered more than 10 million works. Unregistered works generally can't recover statutory damages or attorney's fees. Register scripts, cuts, key art, stills and soundtracks.
- Use technical controls, and document them. The DMCA count depends on an access control that "effectively" restricts access. A real login or token-gated screener is stronger evidence than a CSS overlay. If you share screeners or press kits online, use systems that log access.
- Know who is crawling you. Check your server logs and robots.txt. As the Brave allegations show, some crawlers may be hard to block individually. Many AI crawlers can be blocked by name; keep a record of what you blocked and when.
- Send notices and keep copies. The Times' 2025 and 2026 letters are now central to its willfulness argument. A dated, specific notice to a company you believe is misusing your work can significantly increase what you can recover later.
- Watch the licensing market. These suits are partly about forcing AI companies and data suppliers to pay for licenses. Rights holders with well-organized, registered catalogs will be in the best position to negotiate when that market forms.
IP Feed is tracking both NYT v. Brave and Daily News v. Brave and will report on judge assignment, Brave's response and any motion to dismiss.
This post is editorial commentary based on the CourtListener dockets for The New York Times Co. v. Brave Software, Inc., No. 1:26-cv-09549 (S.D.N.Y., filed Nov. 2, 2026), and Daily News, LP v. Brave Software, Inc., No. 1:26-cv-09554 (S.D.N.Y., filed Nov. 2, 2026), and on reporting by Reuters, The Verge, TheWrap, The Next Web, MLex, IBTimes and PPC Land. It is not legal advice. The allegations have not been proven, Brave has not yet responded, and descriptions of the complaints are as reported in the press. We have not reviewed the complaints themselves.