The New York Times has sued Brave, alleging that the browser and search company copied material from a web-scale collection of more than 30 billion pages and commercialized access to that index for artificial-intelligence services. The claims are allegations in a newly filed complaint; Brave had not filed its response in the supplied reporting.
Brave operates an independent search index and sells query access through the Brave Search API. Its public materials identify Cohere, Mistral, Perplexity and You.com among AI companies using the product. AI assistants can use such an API to retrieve recent information without building and continuously refreshing their own complete index of the web.
The lawsuit focuses partly on how publishers can control collection. Websites commonly place instructions for automated crawlers in a `robots.txt` file. Those rules rely on bots identifying themselves through a user-agent name, which lets a site admit one crawler and reject another. According to Brave's own help material as described in the supplied report, its crawler does not use a distinct Brave user agent. Brave says the choice helps it access sites that permit Google while blocking other search engines.
The Times argues that this approach deprived website owners of a specific mechanism to exclude Brave while remaining visible to other search services. The complaint alleges that Brave gathered pages without identifying itself and then sold access to the resulting index. Whether that conduct violated enforceable rights, and how Brave characterizes its collection and contracts, will be questions for the litigation rather than facts established by the filing itself.
The dispute highlights a structural change in web search. A search index is no longer used only to produce a familiar page of links. It can also become infrastructure for assistants that retrieve documents and synthesize answers. That creates additional customers for indexing services, but it also intensifies long-running arguments over crawler transparency, publisher choice and compensation for material used downstream.
The report also notes uncertainty about which AI products currently rely on Brave. Historical announcements and matching search results have suggested relationships at particular times, but vendor configurations can change. The lawsuit names several companies as Brave customers but does not sue those firms, according to the supplied account.
For website operators, the immediate issue is less whether an ordinary reader uses Brave Search than whether an intermediary's crawler can collect their pages for products they never directly see. The case could test how existing copyright and access rules apply when a search index is packaged as a paid data source for generative AI.
