New York Times Sues Brave: Scrapes Web Pages to Build Index, Then Sells Access to AI Companies

The New York Times and a group of publishing institutions have filed a lawsuit against Brave Software. The complaint targets not only web scraping, but also paywall circumvention.

Per The Verge's report citing Reuters, The New York Times and multiple publishing outlets have initiated legal action against the company that develops the Brave browser, alleging it illegally scrapes content for AI training and allows users to bypass media paywalls.

One of the complaints submitted by the New York Daily News, Chicago Tribune and Orlando Sentinel states that Brave Software monetizes the index built from scraped websites, and provides AI companies with paid access to real-time online content in the "Brave Index". AI companies use this content to train models, answer specific queries, or both.

A web index is the database used by search engines to answer queries. Ordinary users only see a search box, but it is backed by a large number of scraped and organized pages. The core dispute of the lawsuit is whether these pages can be legally scraped, stored, and then sold to AI companies in this way.

Paywall circumvention is another point of conflict. Media outlets claim that Brave lets users bypass paywalls, which effectively makes content that originally required a subscription to read directly accessible for free. This will hurt publishers' subscription revenue and make licensing negotiations far more difficult.

AI copyright disputes are often framed as problems related to model training. This lawsuit pushes another previously overlooked link to the forefront: the web indexing process itself. A browser company that markets itself on privacy is accused of selling indexes of scraped web pages to AI firms, a situation that raises major reputational and legal concerns.

The case is still in litigation, and whether the allegations are valid will be decided by the court. But it at least shows that media outlets are expanding their accountability efforts against AI companies further up the content supply chain, moving from model training to the scraping, indexing and access layers. For teams building AI products, this means compliance of data sources will no longer be limited to the training set stage.

Source: The Verge