New York Times sues Perplexity AI for ‘illegal’ copying of content

The New York Times has launched a major legal challenge against Perplexity AI, accusing the fast growing artificial intelligence startup of illegally copying and distributing millions of its articles to power its generative AI tools. The lawsuit, filed on Friday, represents the most aggressive action yet by a major newsroom seeking to protect its journalism from being used without permission in the AI boom.

According to the Times, Perplexity has been scraping and republishing its reporting on a massive scale, then using that material to train systems capable of generating summaries, answers and long form responses that compete directly with the newspaper’s own content. The complaint argues that Perplexity has built its product on top of the Times’s intellectual property without licensing agreements, payment or attribution, giving it an unfair advantage in the rapidly expanding AI market.

This is not the first confrontation between the two. The Times had previously issued a cease and desist notice to Perplexity, demanding the company stop using its journalism as training data. The lawsuit claims those warnings were ignored, prompting the paper to now seek damages as well as an injunction to block further use of its material.

Perplexity is already facing legal pressure from other publishers. As the demand for high quality datasets intensifies, news organisations have become increasingly vocal about what they describe as widespread misuse of their reporting. Many argue that generative AI systems rely on years of costly journalistic labour, only to produce outputs that draw audiences away from the original sources.

The Times’s lawsuit highlights concerns shared by many in the industry: that without strong protections, AI companies will continue to absorb and repurpose professionally produced reporting, weakening the financial foundations of independent journalism. The paper says Perplexity’s replication of its content is not incidental but systematic, alleging that the startup’s tools “copy, reproduce and distribute” articles in ways that go far beyond fair use.

Perplexity has positioned itself as an up and coming rival in the hyper competitive AI assistant market, rapidly gaining users and attracting investor interest. Its model delivers conversational answers backed by citations and claims to improve accessibility to reliable information. But publishers say those citations often fail to compensate for the underlying issue: that the information being repackaged was never licensed in the first place.

Legal experts expect this case to become a landmark test of how copyright law applies to generative AI, particularly in regard to web scraping, training data and the reproduction of written content. Courts have yet to fully address the extent to which tech companies can ingest copyrighted material in order to build AI systems, and the outcome could shape the future balance between journalism and AI innovation.

As the lawsuit proceeds, pressure is likely to mount on other AI firms to reach licensing agreements with content creators. The Times, one of the world’s most influential newspapers, is signalling that it is prepared to fight hard to protect its reporting from unlicensed use.