NYT sues Brave over Search API, naming Anthropic, Perplexity, Mistral, xAI and Cohere as customers

The New York Times has sued Brave in federal court in Manhattan, targeting the company's Search API as what the complaint calls "an illicit source of unlicensed copyrighted material" — and naming five AI companies, Anthropic, Perplexity, Mistral, xAI and Cohere, as buyers of the product. None of the five AI firms is a defendant. The suit was filed on November 2, 2026, according to TechCrunch, in a report rounded up in a dev.to writeup that also draws on PPC Land, Search Engine Journal and Semafor.

A parallel complaint came the same day from the publishers of the Daily News, the Chicago Tribune and the Orlando Sentinel. Both cases sit in the Southern District of New York, docketed as 26-cv-9549 and 26-cv-9554, and both demand jury trials, according to PPC Land.

What the complaints allege

Each complaint brings claims for direct, contributory and vicarious copyright infringement, anti-circumvention under Section 1201 of the DMCA, and unjust enrichment under New York law, PPC Land reports. The Times describes Brave as a "supplier, dealer, and distributor of stolen copyrighted content" and seeks an injunction plus damages, according to TechCrunch.

The suits reach beyond the API. They also challenge a Brave browser feature the Times says gets around its paywall, and what the publishers describe as Brave's practice of blocking their ads while selling its own. Search Engine Journal reports the Times sent cease-and-desist letters in August 2025 and April 2026, and cites them in court as evidence of willful infringement. The Times has published the complaint itself.

Why the Search API sits at the center

The complaints name four endpoints: News Search, Web Search, LLM Context and Answers. Per PPC Land, the filings describe LLM Context as returning extracted page chunks and source metadata "in a shape and length that satisfies the needs of an LLM," and identify Answers as the service behind Brave's Ask Brave product.

The Times' core argument is a distinction between products: a classic search API returns links and short snippets, while an endpoint engineered to hand page content to a model for reasoning is closer to delivering the article itself. Brave's own pricing page, as of November 2026, lists the Search plan at $5 per 1,000 requests, and Answers at $4 per 1,000 requests plus $5 per million input and output tokens.

The crawler dispute

TechCrunch reports the Times argues Brave built an index of more than 30 billion pages with a crawler it never identified, one that publishers cannot block. Brave's crawler documentation says the bot "does not advertise a differentiated user agent" because the company "must avoid discrimination from websites that allow only Google to crawl them." The same page says Brave will not crawl what Googlebot cannot, and that a noindex directive delists a page.

The practical effect, which the complaints lean on: a publisher that wants Google traffic but not Brave's indexing has no separate robots.txt rule it can write.

Brave's position

Brave did not respond to TechCrunch's request for comment. It has litigated adjacent territory before. Brave sued News Corp first, after receiving a cease-and-desist letter, arguing that indexing website content is something "all search engine operators must do to exist." News Corp then filed its own scraping suit against Brave in July 2026, Semafor reported.

A search engine necessarily copies pages to index them at all. The open question is whether serving extracted page content to AI companies still counts as "search." No court has answered that here, and a verdict is likely years of motions away.

Why it matters

The Times has already sued OpenAI and Perplexity on similar claims, but those cases went after the AI companies themselves. This one targets a supplier further down the stack: the index that feeds models at answer time. If a court accepts that an LLM-shaped retrieval endpoint amounts to distributing copyrighted text, the grounding supply chain behind countless RAG pipelines and agents comes into legal question — and PPC Land reads the requested injunction as broad enough to cover the API endpoints themselves.

Nothing changes for existing API customers today; these are allegations, not findings. But teams building on search APIs now have a concrete reason to verify storage and training rights in their plan terms, log which provider grounded each answer, and keep the retrieval layer behind an interface they can swap quickly.