The New York Times Sues Brave Over AI Scraping and Paywall Bypass: What Is Alleged

The New York Times and a group of other news publishers have sued the company behind the Brave browser. They say Brave scraped their sites without permission, sold that content to AI companies through its search index, and gave its users a tool to skip paywalls. The Verge reported the filings on November 2, 2026, citing Reuters. Everything below is an allegation from a complaint. No court has ruled, and Brave had not replied to requests for comment in the reports we read.

This post sets out what the complaints say, what is confirmed and what is not, how it connects to the News Corp case against Brave from July, and what it means for people who build on search APIs.

NYT v. Brave at a glance

What do the complaints say?

According to The Verge's report, the complaint from the New York Daily News, Chicago Tribune and Orlando Sentinel companies says Brave "monetized its index of scraped websites" in part by giving AI companies paid access to "real-time online content from the Brave Index." The AI companies then use that index to train models, to answer specific user queries, or both.

The same complaint alleges Brave has supplied what it calls "stolen copyrighted content" to "many" AI companies, naming Anthropic, Perplexity, Mistral, xAI and Cohere. The publishers also point at the browser itself. Brave promotes "hundreds of top publishers" as reachable through its "Paywall blocker" feature.

One quoted passage sums up the case: "Brave has never asked for the Publishers' permission to access their websites, much less to copy the material and provide it to users or other AI companies." The publishers add that Brave allows users to bypass paywalls "without their consent" and has "taken other steps that prevent the Publishers from blocking Brave's access." They say Brave never paid them.

The New York Times spokesperson Graham James said the Times 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." He also said Brave ignored "repeated requests by The Times to stop its infringing conduct."

We did not read the complaints themselves. We read the quotes as they appear in The Verge's article, and the dockets are the place to check the full claims.

What did the named AI companies say?

The Verge says Brave, Anthropic, Mistral, SpaceXSI and Cohere did not immediately reply. Perplexity's chief communications officer, Jesse Dwyer, gave a statement mocking the suit: the Times "is overpaying its attorneys and underpaying its journalists." That is a rhetorical reply, not a rebuttal of the factual claims.

Silence in the first hours of a lawsuit means little. It does not confirm anything, and it does not deny anything.

How does this connect to News Corp v. Brave?

This is not the first publisher fight with Brave. On July 21, 2026, News Corp, the parent of The Wall Street Journal and the New York Post, countersued Brave. TheWrap reported the key allegations from the filing:

• Brave "masks its web crawlers such that publishers cannot detect or reliably block them."
• Brave bundles snippets of up to about 250 words, "approximately 5 times traditional search snippets," or its summarizer version of the content, and sells that to enterprise customers, mostly AI companies.
• News Corp calls the product Brave's "Data for AI API" business. Its filing says a "more apt description is theft by a masked intruder."

The dispute began with a February 2025 cease-and-desist letter from News Corp. Brave replied by suing first, arguing that indexing public web content is what every search engine does. News Corp says it then spent more than a year negotiating a licence before Brave "walked away." News Corp asked for a finding of infringement, a permanent injunction, destruction of databases holding the material, and damages.

The new lawsuits use a similar theory. Both stress three points: crawlers that are hard to block, longer-than-usual excerpts, and resale to AI firms. The Times case adds the paywall blocker, which raises a separate question about how a browser feature affects access to paid content.

The legal question: is an index a search engine or an AI data feed?

Search engines have long copied web pages to build an index, and courts in the US have often treated snippet-style display as lawful. The publishers' argument in both cases is that Brave goes beyond that tradition. They describe a line between a "tool of discovery" that points users to the source and a "tool of competitive substitution" that stands in for it, using News Corp's own wording.

Brave's side, from the earlier case, is that the News Corp letter "challenged the most basic function of a search engine." Expect the same defence here: indexing public content is not infringement.

Where a court draws the line is open. Three factors will likely matter, based on the allegations:

1. Crawler transparency. Whether Brave's bots identify themselves and respect blocking rules such as robots.txt. The publishers say they do not. That is a fact question.
2. Excerpt length and substitution. Whether longer snippets and summaries replace a visit to the publisher.
3. Who the customer is. Selling an index feed to AI developers for training and answering is different from showing a result page to a person.

We have followed related fights. Canadian publishers sued Anthropic in Ontario over alleged unlicensed scraping for Claude, and OpenAI has published amicus excerpts arguing fair use in the NYT case. The Brave suits are a step upstream. They target a data supplier that sits between publishers and many AI products.

Why this matters for builders

Many AI products do not crawl the web themselves. They call a search API and receive fresh results, snippets and sometimes full text. If a supplier's index is found to be infringing, the legal risk can move to the customers, or at least to the supply chain. That is the part to watch if you build on a third-party index.

• Ask your search provider about sourcing. How does it crawl, does it honour blocking rules, does it hold licences, and what does its contract say about indemnity?
• Keep records of what you pull. If a question about provenance arrives, you want to know which results you stored, trained on or displayed in full.
• Check how much text you return. The complaints focus on long excerpts. Short snippets with a link carry a different risk from near-verbatim copies.
• Do not assume a feature is neutral. The paywall blocker claim shows a client-side tool can become a legal target.

Brave also runs a private AI chat product. We covered how Brave Leo now encrypts AI chats end to end, which is a separate matter from its search index, but it is the same company and the same push to be an AI-friendly alternative to larger search engines.

What is confirmed and what is not

What to watch next

• Brave's response. A motion to dismiss, a statement, or a settlement would show whether it sees the cases as one fight or several.
• The named AI companies. Whether any of them confirm they buy the Brave Index, and whether they are added as defendants.
• Consolidation with the News Corp case. Overlapping claims about the same product may be handled together.
• Licensing deals. Publishers say they want paid licences. News Corp says it never refused to license. A deal would end the dispute, as the fight often is about price.

We will update this post when Brave responds or a court acts.

Related reading

• Anthropic sued by Canadian news publishers
• OpenAI's fair use amicus excerpts in the NYT case
• Brave Leo end-to-end encrypted AI chat
• Wikimedia's amicus brief on training versus retrieval in the NYT case
• Crawl4AI for Amazon Bedrock Knowledge Bases: preparing web pages for RAG
• Sources: The Verge, TheWrap on News Corp v. Brave

Details are accurate as of November 2, 2026. Lawsuits change quickly; check the court dockets for the current state.