The Real Reason The New York Times Is Suing Perplexity AI (It’s Not About “Theft”)
The headlines today are screaming: “New York Times Sues Perplexity AI for Illegal Copying.” The narrative being pushed by legacy media is simple: Perplexity is a thief, a parasite sucking the lifeblood out of journalism.
But if you look deeper into the lawsuit filed in the U.S. District Court for the Southern District of New York, a different story emerges. This isn’t a simple case of copyright infringement. This is a desperate, strategic attempt by a media monopoly to legislate a new technology out of existence because they failed to build it themselves.
The New York Times (NYT) is not just suing Perplexity AI for “copying.” They are suing to make the act of reading and summarizing illegal for machines. If they win, they won’t just kill Perplexity; they will break the fundamental way the internet works.
At BroadChannel, we analyze the tech, not the PR. Here is the breakdown of why the NYT’s case is legally shaky, technologically regressive, and why Perplexity might actually be the good guy in this fight.
The Flaw in the NYT’s Argument: “RAG” Is Not Copy-Pasting
The core of the NYT’s complaint is that Perplexity uses Retrieval-Augmented Generation (RAG) to “steal” their content. They claim Perplexity bypasses paywalls and serves up their articles for free.
Why this is misleading:
Perplexity does not “copy and paste.” It does what every human researcher does: it reads a source, understands the facts, and writes a new, original summary based on those facts.
- Facts are not copyrightable. The NYT owns the expression of the news (their specific sentences), but they do not own the news itself. If the Times reports “The President signed a bill today,” anyone—including an AI—is legally allowed to restate that fact.
- Perplexity cites its sources. Unlike a plagiarist, Perplexity aggressively links back to the original article. It acts as a discovery engine. The NYT’s argument that this is a “substitute” product ignores the reality that users who want deep, investigative journalism will still click the link. Users who just want the fact were never going to subscribe anyway.
Expert Counter-Point: “The NYT is trying to extend copyright law to cover ‘facts.’ They are arguing that if an AI reads their paper and tells you what happened, that’s theft. By that logic, a morning news anchor reading the headlines from a newspaper is also a thief. It’s a dangerous expansion of copyright that threatens the free flow of information.”
The “Hallucination” Trap: A Weak Legal Pivot
In a bizarre twist, the lawsuit also accuses Perplexity of “hallucinating”—making up fake quotes and attributing them to the Times.
Why this weakens their case:
You cannot sue someone for copying you and for making things up at the same time.
- If Perplexity is hallucinating, it’s not copying.
- If it’s copying, it’s not hallucinating.
By throwing both accusations at the wall, the NYT reveals their true motive: they don’t care about the specific legal theory; they just want to damage Perplexity’s reputation and slow down a competitor. This “reputational harm” argument is a classic tactic used when the copyright claim is weak.
The Real Motive: Killing the Competition
Let’s be honest about what’s happening. The New York Times is a tech company. They sell subscriptions. Perplexity is a superior tech product that solves the user’s problem (getting answers) faster than the NYT’s product (scrolling through ads and 2,000 words of fluff).
- The “Pay-to-Learn” Model: The NYT wants to establish a legal precedent where AI companies must pay a “licensing fee” just to read the internet. This would entrench giants like Google and OpenAI (who can afford to pay) and kill startups like Perplexity.
- Protecting the Ad Model: The “10 blue links” model of search made publishers rich because users had to click and see ads. The “Answer Engine” model kills that. The NYT is suing to protect an obsolete business model, much like the horse-and-buggy industry trying to ban cars.
Why Perplexity Might Win (The “Fair Use” Defense)
Perplexity has a very strong Fair Use defense, based on the precedent of Google Books.
- Transformative Use: Courts have ruled that scanning copyrighted books to create a searchable index (Google Books) is fair use because it creates a new utility. Perplexity is doing the same: transforming scattered articles into a concise, synthesized answer.
- Market Effect: The NYT claims they are losing money. But can they prove it? Is Perplexity stealing subscribers, or is the NYT just failing to attract young readers? Perplexity drives traffic to sources; if the NYT is blocking that traffic with a hard paywall, that is a business decision, not a legal injury.
Conclusion: Innovation on Trial
This lawsuit is not about “protecting journalism.” It is about a legacy gatekeeper trying to toll-booth the future.
If the NYT wins, the internet becomes a closed garden where only the richest companies can build AI. If Perplexity wins, information remains free and accessible, synthesized by the best tools available.
Perplexity isn’t a pirate. It’s a librarian. And suing the librarian for reading the books has never been a winning strategy in the long run.
Frequently Asked Questions (FAQs)
1. Is Perplexity actually stealing articles?
No. Stealing implies taking the original file. Perplexity reads the public web and generates new text summaries based on facts. Copyright protects the exact words, not the information.
2. Does Perplexity bypass the NYT paywall?
Not necessarily. The NYT claims Perplexity accesses “paywalled” content, but often this data is available in public snippets, cached versions, or syndicated feeds. Accessing publicly available data is not “hacking.”
3. Why does the NYT hate AI search?
Because it breaks their ad model. If you get the answer instantly, you don’t visit their site to see banner ads. They view efficiency as a financial threat.
4. Isn’t “Fair Use” only for humans?
No. “Fair Use” applies to the use of the work, regardless of whether a human or a machine performs the action. Google’s bots “read” the entire web every day; that has been legal for 20 years.
5. What is the “Google Books” precedent?
In 2015, courts ruled that Google could scan millions of copyrighted books to create a searchable database without paying authors, because it was a “transformative” tool that helped people find books. Perplexity will likely use this defense.
6. Why is Jeff Bezos involved?
Jeff Bezos is an investor in Perplexity. The media loves to frame this as “Bezos vs. NYT,” but Perplexity is an independent startup. Bezos’s involvement just ensures they have the cash to fight this lawsuit.
7. Can facts be copyrighted?
No. This is a cornerstone of US law. No one owns the fact that a fire happened or a bill was signed. The NYT can only own their specific description of the event.
8. What if Perplexity loses?
It would be catastrophic for the open web. AI models would be banned from learning from the news without paying. This would likely kill most open-source AI projects.
9. Is the NYT being hypocritical?
Many argue yes. The NYT uses AI in its own newsroom for data analysis and personalization. They are suing to stop others from using the same tech they embrace when it benefits them.
10. Does Perplexity cite its sources?
Yes, prominently. Every sentence in Perplexity has a citation number linking to the source. This drives traffic to the publisher, contradicting the claim that it steals traffic.
11. What is “Hot News Misappropriation”?
It’s an old, rarely used legal doctrine the NYT might try to invoke. It prevents competitors from publishing time-sensitive news immediately after a rival gathers it. However, it’s very hard to prove in the internet age.
12. Why is the “Hallucination” claim weak?
It contradicts the copyright claim. You can’t copy something and make it up. By claiming both, the NYT muddies their own argument.
13. Will this go to the Supreme Court?
It’s very possible. This case (along with the OpenAI lawsuits) represents a fundamental constitutional question about AI and the First Amendment.
14. How is this different from Napster?
Napster distributed exact copies of music files. Perplexity distributes newly generated text based on information. It’s the difference between selling a photocopied book (illegal) and writing a book review (legal).
15. Can the NYT prove lost revenue?
This will be the hardest part. They have to prove that a specific person cancelled their subscription because they used Perplexity. That is a very difficult causal link to establish.
16. Is Perplexity blocking the NYT now?
They might. Perplexity could simply blacklist the NYT domain. This would stop the lawsuit but also remove the NYT from the future of search, making them invisible to the next generation of users.
17. What is the “robots.txt” argument?
The NYT says Perplexity ignored “robots.txt” (code that tells bots “do not enter”). However, robots.txt is a polite request, not a legally binding contract. Ignoring it is rude, but not necessarily a crime.
18. Who stands to gain if the NYT wins?
Google and Microsoft. They have the billions needed to pay licensing fees. A NYT victory would destroy the startup ecosystem and hand the AI monopoly to Big Tech.
19. Is this a “Luddite” lawsuit?
Critics say yes. It follows a historical pattern of old industries using courts to stop new technology (like Hollywood suing VCR makers).
20. What should I do as a user?
Keep using Perplexity. This lawsuit will take years. In the meantime, enjoy the most efficient search tool on the planet before the lawyers try to break it.
