The New York Times sued Brave, not an AI lab. The target is the middleman selling your pages.
Four publishers accuse Brave of copying their articles, bypassing paywalls and reselling the material through its Search API to AI companies including Anthropic and Perplexity. The case tests whether search access and AI access are the same permission.
By PANONDA Newsroom

Foto: Abhishek Navlakha · Pexels
What happened
On November 2, the New York Times, the New York Daily News, the Chicago Tribune and the Orlando Sentinel filed separate suits in Manhattan federal court against Brave, the browser company.
The claim is not that Brave built a chatbot on their work. It's that Brave copied their articles, packaged them, and sold access through the Brave Search API to AI companies — including Anthropic and Perplexity, according to the complaints. The publishers also allege Brave's index holds more than 30 billion web pages, that the company prevented publishers from blocking its crawler, that a browser feature circumvented paywalls, and that Brave stripped publisher advertising while serving its own.
Brave has not filed a response yet. None of this is proven.
Why this one is different
Every AI copyright suit you've read about so far named the lab. OpenAI. Microsoft. Anthropic. The theory was: you trained on my archive without paying.
This suit names the supplier — the layer between your site and the model — and that's the part of the pipeline most publishers have never audited.
Think about what you actually control. You can block GPTBot. You can block ClaudeBot. You can turn on Cloudflare's crawler blocking. None of that stops a search company from indexing you for ordinary search and then selling that index to an AI customer. The crawler in your logs says "search engine." The buyer on the other end is a model.
That's the gap the Times is trying to close. If the court buys the theory, publishers get a claim against anyone in the chain, not just the name on the chatbot. If it doesn't, the practical answer is that leaving your page crawlable for search is also leaving it available for resale.
What this is not
It is not a ruling. These are complaints filed two weeks ago. No judge has evaluated the paywall-bypass allegation, the 30-billion-page figure, or whether Brave counts as a distributor rather than an indexing service.
It is not a finding against Anthropic or Perplexity. They are named as customers in the complaints, not as defendants. Nothing public establishes that either company knew how the material was collected.
It is not an EU story. Separately, on November 3, EU Advocate General Maciej Szpunar said chatbots can infringe a publisher's rights when they reproduce an article in recognizable form, especially after bypassing a paywall — but that copyright protects expression, not facts, and that training may fall under the EU text-and-data-mining exception, per Reuters. That opinion is nonbinding and it is European. It will not decide anything in Manhattan. It's useful only because it points at the same operational question: what did you authorize when you left the page up?
It is also not a damages story yet. No dollar figure, no article count, no disclosure of what Brave allegedly earned.
Why the pressure is real anyway
Ziff Davis said its Google traffic fell 37%. That is the backdrop for every licensing fight on this list. Publishers are not litigating as a hobby — the referral business that paid for the archive is shrinking, and the archive is the only asset left to price.
What to do this week
- Pull 30 days of server logs and list every crawler that isn't a named AI bot. Search providers, SEO tools, archive services. That's your exposure map.
- Check whether your robots.txt treats search crawlers and AI crawlers as the same permission. If it does, you have one lever where you need two.
- Read any search API or syndication agreement you signed and look for downstream resale language. Most contracts written before 2023 don't have it.
- Test your paywall against a browser reader mode — Brave's, Safari's, Firefox's. If your content renders clean, the bypass isn't hypothetical.
- If you license to an AI company, add a clause requiring disclosure of where they source your content from third parties.
- Don't rewrite your licensing strategy off a complaint. Wait for Brave's answer. Note the docket and revisit in 60 days.
Sources
Everything above was checked against these pages. Open them and see for yourself — that is why they are here.
Get the stories by email
What changes in AI for people who publish — sourced, and to the point.
No spam. Unsubscribe anytime.