Confused AI Says ‘You Can’t Copyright Facts’ in Defense Against CNN Copyright Suit
Television network CNN is targeting artificial intelligence search engine Perplexity in a copyright infringement lawsuit. As reported by the network’s Brian Stetler, the lawsuit, filed Thursday in New York District Court, accuses the AI company of copying and distributing CNN content, including more than 17,000 CNN stories, videos, photos and other published works.
While this is CNN’s first official lawsuit against an AI company, the network joins other publishers who have sued the San Francisco-based startup, including The New York Times and News Corp. According to the lawsuit, CNN tried to strike a licensing deal with Perplexity, but those negotiations did not result in an agreement. CNN previously entered into a content licensing agreement with Meta last year, where the tech giant compensated the media company for using reporting and content to answer questions with Meta AI.
AI products often scan published news and websites to answer user questions with real-time data, accelerating traffic breakdown and revenue from real sources.
In response to the case, Jesse Dwyer, chief communications officer at Perplexity, told Stetler and other media outlets in a statement: “You can’t copy the facts.” The US government’s Federal Copyright Office says: “Copyright does not protect facts, ideas, programs, or methods of operation, although it may protect the way these things are expressed.”
CNN said in a statement that the multibillion-dollar company should not “steal from the companies that create Perplexity’s original content” and that “commercial employees can and should pay to use it.”
A representative for Perplexity did not immediately respond to a request for comment.
AI copyright suits
Perplexity is one of several companies, including OpenAI and Anthropic, that have been battling news publishers and media giants over copyright claims.
(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging that it infringed Ziff Davis’ copyrights in training and using its AI programs.)
More than 100 such cases have been filed. But different conclusions have been reached about whether training AI models on copyrighted data counts as fair use, said Michael Goodyear, an associate professor at New York Law School. Considerations include how the training takes place, what the AI outputs and whether there is any competitive harm to copyright holders.
“No appellate court has ever examined whether these copyright infringement claims against AI companies are valid,” Goodyear said.
On the CNN issue, he said Perplexity is right that the facts are not protected by copyright, but the way CNN presents the facts may be.
“Even short news articles will qualify for copyright protection under the low bar of originality required,” Goodyear said. “The question becomes whether the thousands of violations reported by CNN are copying entire sections verbatim, whether they are omitting specific words or simply copying unprotected facts.”
AI license agreements
As website congestion has drained billions from publishers’ revenue and caused media layoffs, AI firms are exacerbating the problem. According to a new report from the think tank Open Markets Institute, in the past six months, the rate of AI crawlers bypassing payment walls and blocks has almost quadrupled, rising from 3.3% to 12.9%.
That’s why a number of publishers have signed AI content license agreements with technology companies to monetize content used to train AI systems. Another way out of Perplexity would be to renegotiate the license agreement with CNN. Even if Perplexity has legitimate legal arguments, the license agreement can go from unauthorized singing to a legitimate content partnership.
However, a report by the Open Markets Institute says that when it comes to licensing AI content, news and content creators are locked in a double bind. The same tech giants whose AI tools are starving websites are now the ones securing licensing deals meant to replace that lost ad revenue.
