Key Takeaway: CNN’s copyright infringement lawsuit against Perplexity AI stands apart from prior AI copyright cases because of allegations that Perplexity pursued a licensing deal with CNN, failed to reach an agreement, and then used CNN’s copyrighted content without permission. Those facts could doom a defense of fair use. CNN also included trademark claims, adding another dimension to the case and broadening its potential impact.
CNN has sued Perplexity AI, alleging that, in connection with its AI-powered search engine and other AI products, Perplexity copied more than 17,000 CNN stories, videos, images, and other works without authorization. The suit is pending in the Southern District of New York.
Failed Licensing Attempt May Undercut Fair Use
Amid the growing wave of AI copyright lawsuits, this case stands out because it follows a failed attempt to license. According to CNN, Perplexity negotiated a term sheet with CNN to license CNN content, including to access paywalled CNN content for Perplexity’s Comet Plus users. When Perplexity initially announced the launch of Comet Plus, it identified CNN as a “Comet Plus launch partner.” The parties were unable to reach a long-term agreement, and Perplexity has since edited its blog post to remove CNN. Nonetheless, CNN alleges, Perplexity’s AI products reproduce copyrighted works verbatim, even allowing users to circumvent CNN’s paywalls.
These allegations could prove significant both to a court deciding the legal issues and to a jury evaluating which party acted in bad faith. A defendant that sought a license, recognized the value of the content, and then used it without permission may face a harder time persuading a factfinder that its conduct was lawful. The defense of fair use has loomed large in AI copyright suits. When evaluating the fourth fair-use factor of market harm, some courts have refused to recognize a market for licensing content for use with AI models. Evidence that Perplexity sought a license and then infringed after license negotiations fell through could swing that critical fair-use factor in CNN’s favor.
“You Can’t Copyright Facts” Misses the Point
Responding to the lawsuit, a Perplexity spokesperson stated: “You can’t copyright facts.” That statement fails to address the substance of CNN’s claims. In the suit, CNN identifies multiple articles with registered copyrights, alleges Perplexity copied and reproduced those copyrighted works near verbatim, and provides screenshots to support the allegations.
A Prompt Engineering Defense May Face Limits
In other AI copyright cases, defendants have argued that copyright holders engineered unnatural prompts to force verbatim reproduction of copyrighted works that would not occur during typical use. While Perplexity may try a similar attack, the argument could carry less weight in this case. According to CNN, Perplexity’s browser allows users to employ certain prompts to circumvent CNN’s paywall. The incentive to access paywalled content for free makes it difficult to argue that normal users won’t use those prompts.
Trademark Claims Add Another Dimension to AI Copyright Litigation
CNN has also asserted Lanham Act violations, including claims for trademark infringement and trademark dilution. This is not the first time a news organization has sued Perplexity on such claims. But earlier complaints have alleged harm arising from misattributing content to the news organization, including hallucinated content. In addition to such claims, CNN also alleges Perplexity’s chatbot falsely claimed CNN was a launch partner for Perplexity’s Comet Plus subscription that provides a “premium news bundle,” implying an affiliation between the two companies that does not exist.
What This Case Means for Copyright Holders and AI Companies
CNN’s lawsuit adds to the growing evidence that content owners are prepared to litigate aggressively against AI companies that use copyrighted material without a license. For content owners, the case illustrates how exploring licensing and combining copyright and trademark claims can strengthen enforcement efforts against AI platforms. For both content owners and AI platforms, this is a case to watch: Will the failed attempt to license lead to judicial recognition of a market for licensing content for AI use? The answer could change the trajectory of fair use in AI suits.
Subscribe for News and Updates
