The New York Times filed a copyright infringement lawsuit Friday against Perplexity — the latest clash between traditional media outlets and artificial intelligence (AI) companies over the unauthorized use of journalistic content.
The complaint, filed in the Southern District of New York, alleges Perplexity illegally scrapes Times articles, videos, podcasts, and other materials to generate responses for its artificial intelligence (AI) search engine. According to the lawsuit, the startup produces outputs that closely mirror or directly replicate Times content without permission.
“While we believe in the ethical and responsible use and development of AI, we firmly object to Perplexity’s unlicensed use of our content to develop and promote their products,” Times spokesman Graham James said. “We will continue to work to hold companies accountable that refuse to recognize the value of our work.”
The Times also accused Perplexity of fabricating information and falsely attributing it to the newspaper.
The lawsuit comes one day after the Chicago Tribune filed similar copyright claims against Perplexity, alleging the company scraped and distributed Tribune content without authorization.
Perplexity, founded in 2022, operates an AI search engine that delivers concise answers to user queries. The company has secured more than $1.5 billion in funding from prominent investors including IVP, New Enterprise Associates, and NVIDIA Corp., according to PitchBook data.
Jesse Dwyer, Perplexity’s head of communication, dismissed the legal action as part of a historical pattern. “Publishers have been suing new tech companies for 100 years, starting with radio, TV, the internet, social media and now AI,” he said. “Fortunately it’s never worked, or we’d all be talking about this by telegraph.”
Randy McCarthy, head of an intellectual property law group and a frequent expert source on AI copyright issues, said the case breaks new ground in AI litigation. A general principle emerging from various lawsuits against generative AI companies holds that actions permissible for humans are typically permissible for AI systems, and vice versa, he explained. For instance, just as humans can read books and use that knowledge to create new works without copying directly, courts have generally found that using publicly available copyrighted materials as AI training data does not constitute infringement per se, provided the output is not a direct copy.
“In this case, the NYT is arguing that the summary of materials from the NYT website by Perplexity is actionable on the basis that this denies the NYT ad revenue from accessing that page, as well as other revenues such as subscriptions,” McCarthy said in an email. “There is also a trademark misappropriation claim on the basis that the reader may be confused that the information is in fact coming from the NYT rather than Perplexity.”
McCarthy added that while some of Perplexity’s actions may be lawful, others appear to cross legal boundaries — a pattern common in such cases.
The Perplexity lawsuit represents the Times’ second major legal offensive against AI companies. The newspaper is currently pursuing a 2023 copyright suit against Microsoft Corp. and OpenAI in the same federal court, claiming those companies improperly trained their AI models on Times content.
The legal battles reflect growing tensions between media organizations seeking to protect their intellectual property and AI companies building products on vast datasets that often include copyrighted material. In September, Anthropic agreed to pay $1.5 billion to settle a class action lawsuit with authors who alleged the company illegally downloaded their books from pirated databases — the largest publicly reported copyright recovery to date.
