In a lawsuit filed Thursday in a New York District Court, CNN alleges that Perplexity AI, a San Francisco startup that powers an AI‑driven search engine, has illegally copied and redistributed more than 17,000 of the network’s stories, videos, images and other published works. The network’s legal filing claims the AI service scraped CNN’s material without permission, violating the broadcaster’s copyright and depriving it of revenue.

According to CNN’s statement, the company attempted to negotiate a licensing agreement with Perplexity before resorting to litigation, but talks stalled. The network says a company valued at "tens of billions of dollars" should not "steal from entities that create the original content Perplexity exploits" and insists that commercial operators must pay for the use of such material.

Perplexity’s chief communications officer, Jesse Dwyer, responded that "you can’t copyright facts," echoing the U.S. Copyright Office’s clarification that copyright protects the expression of ideas, not the ideas or facts themselves. The office notes that while facts are unprotected, the specific way they are presented may be eligible for copyright.

Legal scholars note that the dispute hinges on whether Perplexity reproduced whole paragraphs verbatim or merely extracted unprotectable facts. Michael Goodyear, an associate professor at New York Law School, said the low bar for originality means even short news articles often qualify for protection. He added that no appellate courts have yet ruled on the broader question of AI‑training infringement.

The lawsuit adds to a growing tally of more than 100 cases filed by publishers against AI developers, including suits by The New York Times, News Corp and, more recently, Ziff Davis against OpenAI. Companies such as OpenAI and Anthropic face similar challenges as their models learn from vast swaths of online content.

Industry data show that AI crawlers have increasingly bypassed paywalls, with the Open Markets Institute reporting a rise from 3.3% to 12.9% of traffic in the past six months. Publishers argue that the surge in unauthorized scraping is accelerating declines in website traffic and advertising revenue, prompting layoffs across the sector.

In response to the revenue squeeze, several media outlets have entered licensing agreements with tech firms, allowing AI services to use content in exchange for compensation. CNN’s lawsuit suggests that, absent a deal, the network prefers to pursue legal recourse. Analysts say a licensing arrangement could convert the current adversarial stance into a partnership, but the same tech giants that power AI tools also control the licensing negotiations, creating a “double bind” for publishers.

Both sides remain poised for a courtroom battle that could set a precedent for how AI companies train models on copyrighted material. As the case proceeds, the industry will watch closely to see whether the courts uphold the notion that facts are free for all or whether the expression of those facts will receive stronger protection against AI scraping.

This article was written with the assistance of AI.
News Factory SEO helps you automate news content for your site.