“They will only get in trouble if they incorrectly summarize the story and make it defamatory when that wasn’t the case before. “That’s something they’ll actually be at legal risk for, especially if they don’t attribute the original source clearly enough and people can’t easily go to That’s the source to verify,” he says. “If it’s the Perplexity edits that make the story defamatory, then 230 doesn’t cover that, under the body of case law that interprets it.”
In one case noted by WIRED, Perplexity’s chatbot falsely claimed, despite being prominently linked to the original source, that WIRED had reported that a specific California police officer had committed a crime. (“We have been very upfront that the answers will not be 100% accurate all the time, and may lead to hallucinations,” Srinivas said in response to questions about the story we published earlier this week.) “But the core aspect of our mission is to continue improving.” On accuracy and user experience.”)
“If you want to be formal, I think this is a set of claims that would go beyond a proposal to reject a set of theories,” Grimmelman says. “I’m not saying she’ll win in the end, but if the facts support what Police Officer Forbes and Wired — a group of potential plaintiffs — are alleging, they’re the kind of things that, if proven and other facts were bad for the confused, could give rise to liability.”
Not all experts agree with Grimmelmann. Pam Samuelson, a professor of information and law at UC Berkeley, wrote in an email that copyright infringement is about “using someone else’s expression in a way that diminishes the author’s ability to receive appropriate reward for the value of the unauthorized use.” Probably one sentence verbatim Not a violation.
Bhamati Viswanathan, a faculty fellow at New England School of Law, says she’s skeptical that the summary crosses the threshold of substantial similarity normally necessary for an infringement claim to succeed, though she doesn’t think that’s the end of the matter. “It certainly shouldn’t pass the sniff test,” she wrote in an email. “I argue that it should be enough to get your case past the motion to dismiss the threshold — especially in light of all the signs you have of actual things being copied.”
However, she argues that focusing on the narrow technical merits of such claims may not be the right way to think about things, as tech companies can modify their practices to respect the letter of outdated copyright laws while still blatantly violating their intent. She believes that an entirely new legal framework may be necessary to correct market distortions and advance the core goals of US intellectual property law, among them allowing people to benefit financially from original creative work such as journalism so that they are incentivized to produce it. – With benefits to society, in theory.
“There are, in my view, strong arguments to support the intuition that generative AI relies on widespread copyright infringement,” she wrote. “The opening question is, where do we go from there? The bigger long-term question is: How do we ensure the survival of innovators and creative economies? Ironically, AI is teaching us that creativity is more valuable and sought-after than ever. But even as we recognize this, we see The potential to undermine and, ultimately, destroy the ecosystems that enable creators to make a living from their work. This is the dilemma we need to solve, not eventually, but now.