On 2 November The New York Times and, in a twin filing, the New York Daily News, the Chicago Tribune and the Orlando Sentinel sued Brave Software in the Southern District of New York. The complaints are not a generic “AI stole our articles” rant. They target the one independent Western search index that sits behind a large share of AI answer engines, and they treat Brave Search’s API as a wholesale content pipe rather than a ten-blue-links product.
The Times case is 26-cv-9549. The three Tribune papers filed 26-cv-9554 the same day. Both name a single defendant, Brave Software, Inc., and both share the line that Brave acts “both like a burglar that misappropriates the content of publishers, and an AI ‘fence’ that sells stolen content to third parties.”
Three products, one index
The filings describe a pivot. Brave built an independent index — “over 30 billion webpages” on Brave’s own AI page — by crawling without disclosing a distinct bot name. When search alone did not pay, the complaints say, Brave monetised the index three ways: Ask Brave (retrieval-augmented answers inside Brave Search), the Brave Search API sold to other AI firms, and the Brave browser’s optional Paywall blocker plus default ad stripping.
The API evidence is the part SEOs and AI-visibility teams should actually read. In one Times test the LLM Context endpoint returned 1,291 of 1,295 words of an article, verbatim. News Search and Web Search returned more than a third of another piece. Ask Brave, when nudged, reproduced hundreds of words from paywalled Times, Tribune and Daily News stories and offered to “pull up the full article”. That is retrieval, not model memorisation — the same shape of claim Britannica and Merriam-Webster brought against Perplexity last year.
Named as API customers, but not as defendants, are Anthropic, Perplexity, Mistral, xAI and Cohere. Brave’s own pages, as quoted in the complaints, say the API is used by “most of the top 10 AI LLMs” and handles billions of calls a week. We covered Brave’s agent-facing bx402 pay-per-search experiment last week; this litigation is about the older, licensed Search API that already feeds grounding and training pipelines.
The ad count that does not need copying
Count V is the odd one for marketers. Unjust enrichment under New York law, the papers say, “does not depend on any copying”. It alleges Brave blocks publishers’ ads, sells its own, and should disgorge that revenue. Separate from copyright, that puts ad-replacement browsers in a US courtroom as a money claim, not only a product-design debate.
The Times says it warned Brave in letters dated 11 August 2025 and 13 April 2026. Brave has not yet answered either complaint in court, and had not issued a public response on filing day according to contemporaneous reports.
My take
Most publisher AI suits hit the answer engine. These hit the search wholesaler. If an injunction reaches News Search or LLM Context, the citation map inside every AI product that buys Brave’s index moves. For anyone auditing brand visibility in Ask Brave, Perplexity or similar RAG stacks, annotate 2 November: the legal risk now sits on the independent index itself, not only on the chat UI. Watch for Brave’s reply, any interim crawl or endpoint changes, and whether other publishers join. The case numbers above are the primary sources; PPC Land’s read of both complaints is the clearest secondary walkthrough so far.