AILSA CHANG, HOST:
The New York Times and Chicago Tribune have sued the artificial intelligence company Perplexity AI. These were separate cases last week for copyright infringement. And they're the latest in a wave of lawsuits against AI firms by media companies trying to protect their business model. To explain all of this, we've got NPR's John Ruwitch on the line. Hi, John.
JOHN RUWITCH, BYLINE: Hey, Ailsa.
CHANG: OK, so explain what exactly is at the heart of these latest cases.
RUWITCH: Yeah, they're really about who can reuse content that's been published on the internet by news companies and under what circumstances. Perplexity is one of the most popular AI chatbots, like ChatGPT or Google's Gemini. These chatbots are trained on data taken from the internet. And so you ask it a question, it gives you an AI-enabled answer. In a nutshell, what these news companies are arguing, though, is they're accusing Perplexity of using their content without permission to create some of those answers. Steven Lieberman is a lawyer with the firm Rothwell Figg. He's representing The New York Times and the Chicago Tribune. He says Perplexity should pay for the content, just like other kinds of companies have done for years and years.
STEVEN LIEBERMAN: So for example, with television, one of the most popular early shows was CBS' "I Love Lucy." But CBS didn't take "I Love Lucy's" content without paying Lucille Ball and Desi Arnaz. They paid them for the content.
CHANG: And how has Perplexity responded to this?
RUWITCH: Well, Perplexity kind of shrugged it off in a statement that they sent to journalists. The company's head of communications, Jesse Dwyer, said that publishers have been suing new tech companies for a hundred years. And, quote, "fortunately, it's never worked, or we'd be talking about this by telegraph."
CHANG: (Laughter).
RUWITCH: It was kind of snarky.
CHANG: Yeah (laughter).
RUWITCH: But the crux of these disputes, yeah, the crux is generally, you know, this question of what is considered or what should be considered fair use and whether or not AI changes things. Fair use is this legal concept that lets copyrighted works be used by others without the owner's consent for, like, educational purposes or research purposes, but with conditions. So it has to be transformed into something new.
CHANG: Right.
RUWITCH: I spoke to Sarah Kreps about this. She leads the Tech Policy Institute at Cornell University. She says AI companies are really testing the boundaries of fair use by hoovering up every piece of data and content that they can get their hands on.
SARAH KREPS: It's not entirely clear with a new technology what those boundaries are. And so it's not surprising that they would push and see what those limits are.
RUWITCH: That's because, basically, competition in AI is cutthroat. AI models are built on data. The more data, the better the model will be.
CHANG: OK, well, John, with all these AI stories, I feel like we keep hearing AI is a threat to this or it's going to destroy that. So just explain really specifically here, why is AI so threatening to these media companies?
RUWITCH: For media companies, it's almost existential, right? They are in the business of creating news content and selling it. And analysts say actually grim lessons of the past are helping to animate some of these cases. You will remember, Ailsa, that the internet radically changed the news industry, right?
CHANG: Oh, yeah.
RUWITCH: Put a lot of newspapers out of business by undercutting ad revenues, giving customers new ways to get the news for free. News companies don't want something like that to happen again. Klaudia Jazwinska is a researcher and journalist at the Tow Center for Digital Journalism at Columbia University.
KLAUDIA JAZWINSKA: I think a lot of publishers feel like they've been burned in the past. And they don't exactly feel like they have a reason to trust AI companies or tech platforms.
RUWITCH: She and her colleagues actually launched an online tracker last week. And they've tallied around 20 lawsuits so far - she says there could be more - and over a hundred deals where media companies are getting paid for their content through licensing or revenue sharing agreements.
CHANG: Will we see more deals, you think?
RUWITCH: We might. Steven Lieberman, the lawyer for the newspapers, says that that's where things should head. The two sides need to ultimately sit down and hammer out licensing deals. But these, you know, will take time. The cases are grinding their way through the courts. The original, the first one, was launched about two years ago. That was New York Times v. OpenAI. And there's no end in sight to that case right now.
CHANG: That is NPR's John Ruwitch in Silicon Valley. Thank you, John.
RUWITCH: You're welcome. Transcript provided by NPR, Copyright NPR.
NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.
