S3DAT NEWS
· 16 stories RSS s3dat.com

Publishers sue Brave over AI search

The New York Times and three other newspapers sued Brave on November 2 over its use of their journalism. The cases reach the search infrastructure that supplies material to AI applications.

Spec sheet

Defendant
Brave Software
Times case
26-cv-9549
Other publishers’ case
26-cv-9554
Court
Southern District of New York

A challenge to retrieval and reuse

The Times alleges that Brave copied articles into its index and supplied unlicensed text through Ask Brave and its search API. Its complaint also challenges paywall circumvention and the substitution of advertising. It seeks damages, restitution and a permanent injunction. In one cited test, the Times says the LLM Context endpoint returned 1,291 of an article’s 1,295 words. The publisher also describes cease-and-desist letters sent in August 2025 and April 2026. It alleges Brave ignored both.

A second complaint provides different examples

The Daily News, Chicago Tribune and Orlando Sentinel filed their own action against Brave the same day. Their complaint describes a conversation in which Ask Brave offered to retrieve more Daily News articles, then produced nearly 600 words about New York horse carriages. Another example concerns a Tribune report about data-center tax breaks near O’Hare: the publishers say the answer reproduced more than 1,200 words. The filing includes screenshots and side-by-side comparisons, with a fuller comparison for the Tribune example in an exhibit. The same complaint describes an Orlando Sentinel example in which the answer combined verbatim passages and close paraphrases.

How Brave described the product

Brave’s February 12 product announcement described LLM Context as a way to give a model relevant material from the web. The service searches its independent index, extracts content from selected pages and ranks passages for the requested query. Its supported extraction includes tables, structured data, code, forum discussions and YouTube captions. Developers can set a token budget and adjust how many URLs contribute. Brave also described Goggles controls that filter or change the ranking of sources by domain or URL pattern. This puts the choice of retrieved material between the publisher’s page and the application’s own model.

Brave’s earlier legal position

Brave argued for fair use in a separate complaint against News Corp entities filed on March 12, 2025. It argued indexing, summarizing and producing limited excerpts served a different purpose from publishing the underlying articles. Its argument emphasized the factual nature of news, the limited material it said it used and links that send readers to publishers. Brave sought declarations of non-infringement, copyright misuse and no breach of contract. It also disputed that it had agreed to the publishers’ website terms. That filing sets out the company’s earlier legal theory in a different dispute; it is not an answer to the November 2 complaints.

Claim check

  • Unclear These complaints have not established copyright infringement or unlawful paywall circumvention. [1]
  • Company claim The cited output examples are publisher-reported tests, not an independently measured rate of article reproduction. [2]

Sources