New York Times Sues Brave Over AI Search API

On Monday, November 2, the New York Times sued Brave Software in federal court in Manhattan. The same day, the publishers of the New York Daily News, the Chicago Tribune and the Orlando Sentinel filed their own suit in the same court. I read the parts of the Times complaint that describe Brave’s products, and the browser is the third of three products it goes after.

What is the Times actually accusing Brave of?

Three things, in the complaint’s order. Brave built a search index by copying pages from across the web. It then monetized that index, first through an AI answer feature and then through a paid API. And its browser has a setting that gets readers past paywalls.

Brave product	What the Times complaint says it does	Brave’s own wording, as quoted in the complaint
Ask Brave	Gives AI answers that often include verbatim text and detailed summaries of articles	Brave calls the method “AI Grounding”
Brave Index and Search API	Sells real-time copies of scraped pages to AI companies for training and retrieval	“ready-to-use content in a single call”
Paywall blocker (browser setting)	Goes behind paywalls and hands the reader the full article	Promoted as letting users “Bypass Paywalls”

The complaint also says Brave’s browser blocks the Times’s ads by default and shows Brave Ads instead, which is where the unjust enrichment count comes from.

Which AI companies are named?

Five named companies. The Times complaint calls Brave a supplier to AI companies including Anthropic, Perplexity, Mistral, xAI and Cohere, and the Daily News complaint names the same five, according to The Next Web. The Times complaint also quotes a Brave blog post saying “Anthropic used Brave’s Search API to train their latest model, Claude 2, on high-quality data to help optimize Claude’s outputs for retrieval augmented generation (RAG) queries.” None of the five is named as a defendant in the Times suit. Brave is, and the complaint calls the AI companies “knowing receivers” of the content. My read: that wording leaves room for the publishers to go after the AI companies later.

Why does the crawler matter?

Websites block bots by name in robots.txt. The Times says Brave has never disclosed the name of its crawler, so there is nothing for a publisher to block. By Brave’s own count, quoted in the complaint, the index holds more than 30 billion pages.

My read: that’s why the new shared opt-out signal from the big labs, which we covered on October 31, wouldn’t help here. A signal only works if the crawler announces itself.

What did the Times do before suing?

It says it wrote twice. A letter to Brave CEO Brendan Eich on August 11, 2025, asked Brave to stop and remove Times content from its products. A second letter from outside counsel on April 13, 2026, said continued use showed “reckless disregard” and could mean statutory damages of up to $150,000 per infringement. According to the complaint, Brave answered neither. The Times uses that silence to argue the infringement was willful, which raises the damages ceiling.

What does it want?

The Times lists five counts: direct, contributory and vicarious copyright infringement, circumvention of technical protections under the Digital Millennium Copyright Act, and unjust enrichment under New York law. It asks for statutory and actual damages, Brave’s profits, a permanent injunction and a jury trial. The Daily News papers seek similar relief, and TheWrap reports their complaint says Ask Brave produced more than 1,200 words from one Chicago Tribune article.

What does Brave say?

Nothing I could find yet. TheWrap reported that Brave didn’t immediately respond to its request for comment. Brave’s earlier position is on record from a separate fight. In March 2025 it sued News Corp, asking a court to declare that scraping, indexing and AI-generated summaries are fair use, and News Corp countersued in July 2026, per TheWrap. The Register quoted Brave’s filing saying “decades of legal precedent” show indexing for a search engine is fair use.

My read

Brave’s fair-use defense is strongest for what search engines have always done: index pages and show a link with a few words. The Times complaint is written to stay away from that ground. It points at the API endpoint that, in Brave’s own marketing, returns page content “without need for further scraping.” Whether that is still search, or a text feed for model builders, is the question a judge will have to answer, and I don’t think anyone can call it yet.

It also connects to the one appeals ruling we have. The Third Circuit held in September that training on Westlaw headnotes wasn’t fair use, but it limited the holding to a tool that returns existing text, as we reported then. A search API that returns existing text sits closer to that case than a chatbot does. That’s an inference from two documents, not a prediction. The court that decides this one isn’t bound by the Third Circuit, and a complaint states only one side.

For now, watch for Brave’s first filing in response. It may show whether the company defends the whole business as search or separates the API from the rest.

Note on sources

I read the Times complaint’s opening, its product descriptions, the two cease-and-desist letters and its counts. I haven’t read the Daily News complaint in full; what’s said about it here comes from The Verge, TheWrap and The Next Web, named inline. I didn’t open the Brave web pages that the complaints quote, so those quotes are as the Times complaint prints them. I also didn’t open the News Corp case filings or The Register’s full article beyond what is quoted, so that background is as those outlets report it. The Third Circuit point rests on our own earlier post, linked above. All claims about Brave’s conduct are allegations in unproven complaints.

The New York Times sued Brave over a paywall setting, and over the search index Brave sells to AI labs.