«Burglar and AI fence»: American publishers sue browser operator Brave
With new copyright lawsuits against Brave Software, leading American media companies are trying out a new legal strategy. They’re specifically holding the operators of web crawlers accountable for providing the raw data used to train AI models.
Several American newspaper publishers, including the New York Times (NYT), the Chicago Tribune, the Daily News and the Orlando Sentinel, are suing Brave Software – a browser and search engine operator. It’s alleged to have committed massive copyright infringements by selling the newspapers’ copyrighted journalistic works to AI companies such as Anthropic, Perplexity and xAI. According to the publishers, this also includes their paywall-protected content.
Two complaints, largely structured in the same way, have been filed with the United States District Court for the Southern District of New York (NYT, other publishers).
Publishers vs. privacy browser
Brave Software was co-founded in 2015 by Brendan Eich, the creator of JavaScript. This business operates the privacy-focused browser Brave and the search engine Brave Search. According to the company’s own figures, the browser reached the milestone of more than 100 million monthly active users in September 2025. It currently has a global market share of 0.4 per cent.
Brave markets its browser as being particularly privacy-friendly, blocking ads, cookie banners and tracking software. A built-in VPN client for premium subscribers encrypts users’ connections and hides their IP addresses. A function called Global Privacy Control is designed to prevent the sharing and sale of personal data.
On the other hand, the browser came under fire in 2020 because it secretly appended affiliate codes for crypto exchanges to the URLs users visited in the address bar, fraudulently earning commission](https://davidgerard.co.uk/blockchain/2020/06/06/the-brave-web-browser-is-hijacking-links-and-inserting-affiliate-codes/).
In the introduction to their complaints, the publishers accuse Brave of acting both as a burglar that unlawfully appropriates publishers’ content and as an AI fence that sells stolen data to third parties. The lawsuit centres around Brave’s search index, which the company built itself. While many search engines rely on search indexes from other suppliers, such as Google, Brave runs its own index. According to the complaint, the database contains more than 30 billion web pages – and Brave Software is alleged to have sold this data to AI companies to train their AI models.
Many publishers, including the plaintiffs, grant search engine crawlers access to content that’s normally only available to regular visitors behind a paywall. This means the pages appear in search results, but when users click on them, they’re prompted to pay. The plaintiffs accuse Brave of using this access for its own purposes. They themselves have no way to specifically block Brave’s crawlers from accessing paywall-protected content.
At the same time, Brave offers its users several ways to access protected content for free. Specifically, the publishers accuse Brave of offering functions that can be used to bypass paywalls. These include AI tools allowing users to specifically request content from protected websites. The tool generates summaries or even almost the complete text of articles, as stated on page 30 of the NYT’s complaint.
What are the publishers demanding?
In their complaints, the publishers cite several possible demands. These include, for example, the maximum statutory damages provided for under US copyright law – up to 150,000 US dollars per work intentionally infringed. The NYT puts this number at more than ten million, while the complaint filed by the other publishers lists more than two million registered works in total. As such, a ballpark figure for the compensation would – in theory – add up to trillions. Amounts of this magnitude represent the maximum theoretical limit.
What’s more, the publishers are demanding that Brave Software repay the profits generated through scraping and selling the content. The complaints cite a figure of more than 100 million US dollars for this. In major US copyright litigation cases, the amounts actually awarded are often set significantly lower or decided by out-of-court settlements.
And that’s not all – compensation under the US DMCA (Digital Millennium Copyright Act) is also possible. Plus, publishers are demanding compensation for lost profits resulting from the bypassing of paywalls.
Do the lawsuits stand a chance of success?
The NYT filed a similar lawsuit back in late 2023. It was directed specifically at AI developers OpenAI and Microsoft. The allegation is that the companies are using copyrighted works without permission to train their AI models. So far, the key copyright allegations have withstood the defendants’ attempts to have the lawsuit dismissed. But a final ruling and the possibility of compensation have still not been decided.
AI companies cite the US fair use doctrine, which permits the citation and use of copyrighted data under certain circumstances. The question of whether AI training is covered by this still hasn’t been settled.
In this particular case, it’s not the AI companies and their use of data that are in the crosshairs, but rather the infrastructure that provides the data. The lawsuits revolve around traditional copyright law instead of new technologies related to AI models and their training. If the court finds that Brave Software systematically copies protected content and markets it commercially to third parties, a fair use defence is likely to become more difficult.
The lawsuits could actually have a greater chance of success for the publishers than accusing the AI companies of using content for their AI training. This could set a precedent for the extent to which search engines, web crawlers and data providers are permitted to use and share copyrighted content for AI applications.
Brave has not yet publicly commented on the allegations. However, the company’s likely to argue that crawling and indexing websites has been the standard way that search engines have worked for decades.
Feels just as comfortable in front of a gaming PC as she does in a hammock in the garden. Likes the Roman Empire, container ships and science fiction books. Focuses mostly on unearthing news stories about IT and smart products.
Interesting facts about products, behind-the-scenes looks at manufacturers and deep-dives on interesting people.
Show all
