New York Times Sues Brave Browser: AI Content Scraping Copyright Battle Escalates

Background: Privacy Browser Accused of Scraping News Content

According to RSS sources, The New York Times is suing privacy-focused browser company Brave, accusing it of scraping the media outlet's copyrighted content and "illegally distributing" it to its own customers. The complaint specifically names Anthropic and Perplexity—two AI companies that indirectly obtained New York Times content through Brave's search or data services. This lawsuit brings the copyright conflict between browsers, AI companies, and traditional media to a new climax.

Brave has always marketed itself as "privacy-first," with its Brave Search engine featuring an independent index and a data service for AI developers. The New York Times argues that Brave scraped news content without authorization and provided it to third parties as training data or search results, constituting copyright infringement. Brave has not yet publicly responded, but this case could become a landmark ruling on content licensing rules in the AI era.

The Core Dispute: Where Is the Boundary Between Scraping and Distribution?

The core of this lawsuit lies in the legal characterization of two actions: "scraping" and "distribution." The New York Times claims that Brave not only crawled its website content but also integrated it into its own products and even sold it to AI companies like Anthropic and Perplexity. This effectively turns traditional media's original content into a "free lunch" for AI companies.

From a technical perspective, Brave's search index does need to crawl web pages, but search engines typically follow the robots.txt protocol and only provide snippets and links. However, when scraped content is used to train large models or directly provided to AI products as a knowledge base, the nature changes. The New York Times previously sued OpenAI and Microsoft, and now targeting Brave shows that content owners are expanding their scope—from model developers to data intermediaries.

The robots.txt protocol is a self-regulatory convention between websites and web crawlers. Websites declare in this file which areas crawlers are allowed or forbidden to scrape, and crawlers such as search engines typically comply. However, this protocol does not have legal force; violating it mainly raises industry ethics and contractual disputes rather than directly constituting infringement. In copyright lawsuits, defendants often use "complied with robots.txt" as a defense, but plaintiffs typically argue that even if the protocol was followed, subsequent content use (such as AI training or resale) may still exceed the authorized scope. Therefore, compliance with robots.txt often serves as a reference factor in determining whether scraping was "in good faith," but it is not a decisive legal basis.

Fair Use is an important defense principle in U.S. copyright law that allows unauthorized use of copyrighted works in specific circumstances, such as for commentary, news reporting, teaching, or research. Determining fair use typically considers four factors: the purpose and character of the use (whether it is transformative, whether it is commercial), the nature of the original work, the amount and substantiality of the portion used, and the effect on the potential market for the original work. In the context of AI scraping and training, defendants often claim their use is "transformative," while content owners emphasize market erosion. In the New York Times v. Brave case, Brave will likely invoke the fair use defense, but how the court weighs these four factors—especially the impact on the news content market—will be key to the precedent.

Impact on the AI Industry: Data Supply Chains Face Compliance Scrutiny

If the New York Times wins, the AI industry's data supply chain will face stricter compliance scrutiny. Many AI companies currently rely on third-party data providers for training corpora, and companies like Brave are exactly the intermediary link. Once a court rules that scraping and reselling news content is illegal, AI companies may need to reassess the legality of their data sources and may even be forced to sign direct licensing agreements with content owners.

On the other hand, Brave's privacy browser business could also be affected. Its search service relies on its own index; if restricted from scraping news websites, the quality and coverage of search results would decline. For users, this could mean reduced functionality in the privacy browser or the need to pay for content. From an industry trend perspective, content licensing is becoming a new barrier to AI competition—media outlets with high-quality copyrights will gain more bargaining power, while companies relying on public scraping face legal risks.

Data supply chain in the AI industry refers to the complete chain from data collection, cleaning, and annotation to final use in model training or inference. Common participants include crawler operators that directly scrape web pages, data intermediaries that provide structured data or APIs (such as Brave Search API), and large model companies that purchase data for training. Since many AI companies do not directly scrape the entire web but rely on third-party data services, compliance risks in the supply chain are passed down layer by layer. Once upstream data scraping is deemed infringing, downstream users may also face joint liability or be forced to change data sources. This is prompting AI companies to increasingly prioritize authorization proof and compliance audits for data sources to reduce legal risks.

Multi-Source Perspectives: The Game Among Parties in the Copyright Battle

According to RSS sources, the New York Times lawsuit also names Anthropic and Perplexity, indicating that content owners believe these two AI companies indirectly obtained infringing content through Brave. However, there is no information suggesting that Anthropic or Perplexity have been named as defendants; they may simply be mentioned as "customers." This reflects the unclear division of responsibility between AI companies and data intermediaries: AI companies can argue they legally purchased data from Brave, while Brave may claim its scraping constitutes fair use.

Interestingly, Brave previously launched "Brave Search API" and "data sales" businesses, providing search index data to AI companies. This business model itself sits in a copyright gray area. The New York Times choosing to sue Brave rather than directly suing Anthropic may be to cut off the data supply chain at the source, or to first establish a precedent favorable to content owners. Regardless of the outcome, this lawsuit will prompt the AI industry to rethink the ethical and legal framework for data acquisition.

Search index is a database built by search engines after parsing, deduplicating, and structuring scraped web page content, used to quickly respond to user queries. Unlike traditional search engines that only return snippets and links, companies like Brave may sell index data in API form to AI companies for model training or enhancing generation. This "data sales" model blurs the line between search services and data distribution: if the index contains copyrighted news full texts or large excerpts, selling index data could constitute reproduction and distribution of the original works. The New York Times lawsuit targets this very link, arguing that Brave's index construction and resale go beyond the scope of fair use.

Conclusion: A Signal of Content Value Returning

The New York Times v. Brave case is essentially another attempt by traditional media to defend content value in the AI era. From suing OpenAI to suing Brave, content owners are building a complete chain of rights protection: whoever scrapes, whoever distributes, and whoever uses may be held accountable. For AI practitioners, the era of relying on public scraped data is ending, and compliant licensing will be key to competition in the next phase. Brave's fate is not only about the survival of a browser company but will also define the rule boundaries of the AI data supply chain.