The AI infringement hits keep coming, reports Blake Brittain of Reuters:

Perplexity AI is the latest artificial intelligence company to be hit with a lawsuit by copyright holders alleging infringement after Encyclopedia Britannica and Merriam-Webster accused it of misusing their content in its “answer engine” for internet searches.

The reference companies alleged in New York federal court on Wednesday that Perplexity unlawfully copied their material and diminished their revenue by redirecting their web traffic to its AI-generated summaries.

Elissa Welle at The Verge:

In the lawsuit, the companies allege that Perplexity’s “answer engine” scrapes their websites, steals their internet traffic, and plagiarizes their copyrighted material. Britannica also alleges trademark infringement when Perplexity attaches the two companies’ names to hallucinated or incomplete content.

The 55-page filing (in which the company delightfully spells its name Encyclopædia) is filled with confusing, sometimes circular contentions. For example,

7. Perplexity’s so-called “answer engine” eliminates users’ clicks on Plaintiffs’ and other web publishers’ websites—and, in turn, starves web publishers of revenue […]

That ship has long sailed. People want answers, not just links. Is this much different than me reading an Encyclopedia Britannica article and then summarizing it in a post here? Am I stealing their content and revenue if my page ranks higher than Britannica’s?

That paragraph continues:

[…] To build its substitute product, Perplexity engages in massive copying of Plaintiffs’ and other web publishers’ protected content without authorization or remuneration.

Ah, there’s the rub! Wholesale copying of content is certainly not acceptable. If that’s the complete complaint, consider me convinced.

Welle’s Verge article also notes:

Perplexity plagiarized Merriam-Webster’s definition of the word plagiarize, the lawsuit alleges.

Plagiarizing “plagiarize” is funny. But after reading the actual allegation, I found it—to be kind—rather specious:

74. For example, when a user asked Perplexity, “How does Merriam Webster define plagiarize,” Perplexity spit back the exact definition of the term from Merriam Webster, which is identical to the definition in the print Merriam-Webster Collegiate Dictionary.

What is it they expect to happen here? Should Perplexity not answer? Link to the Merriam-Webster site?

If you asked me this question, and I pulled out a copy of the Merriam-Webster Collegiate Dictionary and read the definition to you, am I in violation of copyright? Should I be expected to open the dictionary to the definition and slide the book across the table for you to read yourself?

At best it’s a leading question. At worst, it’s contrived—and the allegation contains several such examples.

What if Perplexity returned a different answer? Well, they’d then be accused of providing inaccurate information while attaching Merriam-Webster’s name:

93. […] Perplexity sometimes generates hallucinations in its outputs and attributes that text to Plaintiffs using Plaintiffs’ trademarks. Perplexity’s use of these marks alongside hallucinations is likely to cause dilution by blurring and/or tarnishing Plaintiffs’ famous marks. In addition, Perplexity’s use of these marks alongside hallucinations constitutes false designations of origin and confuses and deceives Perplexity users into believing (falsely) that the hallucinations are associated with, sponsored by, or approved by Plaintiffs.

In addition to complaining about exact reproductions, and “hallucinations” with their name on it, Britannica also complains that Perplexity’s answers aren’t exactly what Britannica wrote, actually:

94. It omits content from Britannica’s articles that it purports are exact reproductions of such articles.

So Perplexity is wrong for reproducing content exactly, inexactly, and inaccurately. In other words, for reproducing the content at all. Got it.

What’s unclear is whether Britannica would allow any of this behavior if Perplexity were to pay a licensing fee, because it seems Perplexity was happy to do so:

86. In May 2025, Perplexity reached out to Plaintiffs to discuss a potential partnership. The parties had an initial phone call on May 9, 2025 during which Perplexity did not provide any material information regarding the requested partnership, financial or otherwise. After the call, Perplexity requested multiple times that the parties negotiate a non-disclosure agreement before engaging in further substantive discussions. Neither Plaintiff ultimately executed a non-disclosure agreement or participated in any partnerships with Perplexity.

I’m all for Perplexity compensating Encyclopedia Britannica to gain access to its valuable content. And stealing content should be properly punished. Yet this lawsuit is filled with such scattershot arguments, it comes across as more a negotiation ploy than a legitimate complaint.