The New York Times Sues Brave, Calling the Browser an AI "Fence" for a 30-Billion-Page Scraped Index

The New York Times sued Brave Software on Monday, alleging that the privacy-focused browser scraped the publisher's copyrighted material and passed it on to AI companies the complaint names as Anthropic, Perplexity, Mistral, xAI and Cohere.

The suit was filed in federal court in Manhattan. In a separate action the same day, the Daily News, the Chicago Tribune and the Orlando Sentinel filed a similar complaint. The Times has already sued OpenAI and Perplexity over related claims; this filing extends the argument to the layer beneath the models.

The language is unusually hostile for a copyright complaint. The Times describes Brave as "a supplier, dealer, and distributor of stolen copyrighted content," and says the company "acts both like a burglar that misappropriates the content of publishers, and an AI 'fence' that sells stolen content to third parties."

The factual core is an index. According to the complaint, Brave assembled an index of more than 30 billion webpages using an unidentified crawler, which the Times says left publishers unable to block it through ordinary means. The suit also targets a Brave feature that bypasses paywalls, and the company's practice of blocking publishers' advertisements while selling its own inventory against the same pages.

That combination — an unidentifiable crawler, paywall circumvention and ad substitution — is what turns a familiar scraping dispute into something closer to a web-integrity case. Publishers can negotiate with a named AI lab, and several have. It is considerably harder to negotiate with a browser that presents itself as the user's agent while running a commercial content index behind it.

Brave, founded in 2015, has raised roughly $250 million to date according to private-market trackers. The Times is seeking an injunction and damages. Brave did not immediately respond to a request for comment when the suit became public.

The case lands at an awkward moment for the licensing market. AI companies spent the past year signing content deals and publishing opt-out standards precisely to reduce this kind of exposure, and an intermediary that supplies the same content without a licence undermines the price of every deal signed above it. If the Times prevails on the fence theory, liability stops running only through model developers and starts running through the infrastructure that feeds them.