CNN sues Perplexity AI over "verbatim" copies of articles

CNN is suing AI company Perplexity claiming its software generates verbatim copies of its news stories. It's the latest example of traditional publishers targeting AI-powered language models surreptitiously trained on their copyrighted material. Files in the Southern District of New York, the federal lawsuit indicates that the two companies tried to negotiate a content deal but failed to close.

The lawsuit claims that "Perplexity unlawfully crawls, scrapes, copies, anddistributes CNN's content from CNN Digital Platforms and third-party platforms," counting 17,000 examples, and then "generate outputs that are identical or substantially similar to CNN's content."

CNN, quoting its own spokesperson anonymously:

"CNN's lawsuit stands for the proposition that Perplexity, a company valued at tens of billions of dollars, should not be able to steal from entities that create the original content Perplexity exploits," a CNN spokesperson said in a statement. "The public rely on high quality news journalism reported by human beings to understand their world, which is frequently dangerous and expensive to produce. Commercial operators can and must pay to make use of it."

"You can't copyright facts," Perplexity's chief communications officer Jesse Dwyer responded, though he did not address the lawsuit's claims of reproducing works verbatim.

As an illustrative example, when a user asks Perplexity's chatbot to, inter alia,"[d]eeply analyze an article from a single reputable source to tell me what's been reported aboutMarco Rubio dropping one of his job titles," Perplexity provides verbatim portions from the CNN article "Marco Rubio drops one of his job titles, while still juggling several other hats."

Perplexity seems to have attracted a lot of negative legal attention; it's also been sued by The New York Times, the Chicago Tribune, Yomiuri Shimbun, and Encyclopedia Britannica. CNN reports that it struck deals with Gannett, TIME, Le Monde and Der Spiegel.

Previously:
• Britannica sues OpenAI for scraping nearly 100,000 articles
• Musk's AI sued after generating explicit images of his co-parent
• RIAA sues AI music startups over copyright infringement claims