“The crawler, which does not contain a unique user-agent string, respects robots.txt.”
Brave sells a Search API tier that lets buyers “Cache/store data to train AI models,” with snippets running 150 to 260 words per result. Its crawler hits more than 40 million pages a day and doesn’t identify itself. Respecting robots.txt means nothing when you can’t name the bot you want to block. The privacy browser found a way to monetize everyone else’s content.