The New York Times has filed a lawsuit against AI search startup Perplexity, accusing the company of copyright infringement. This marks the second such legal action by The Times against an AI company, following an earlier suit against OpenAI and Microsoft.

The core of The Times’ complaint is that Perplexity's commercial products directly substitute for the outlet's content, using it without proper permission or remuneration. The lawsuit highlights that Perplexity's retrieval-augmented generation (RAG) products, including its chatbots and Comet browser AI assistant, crawl websites and databases to generate responses for users. These products often provide verbatim or near-verbatim reproductions, summaries, or abridgments of copyrighted material, including content behind The Times’ paywall. Additionally, The Times alleges that Perplexity's search engine has “hallucinated” information and falsely attributed it to the news organization, resulting in potential brand damage.

This legal action is part of a broader, years-long strategy by publishers to navigate the evolving AI landscape. Recognizing the inevitability of AI integration, media outlets are using lawsuits strategically to compel AI companies to formally license content, ensuring creators are compensated and the economic viability of original journalism is maintained.

Perplexity has attempted to address these compensation demands through initiatives such as a Publishers’ Program, offering a share of ad revenue to participating outlets including Gannett and the Los Angeles Times. It also launched Comet Plus, allocating 80% of its monthly fee to publishers, and secured a multi-year licensing deal with Getty Images. Despite these measures, The Times maintains that Perplexity’s unlicensed use of its content remains unacceptable, a sentiment echoed by spokesperson Graham James.

The New York Times is not alone in raising concerns. Several other media entities have challenged Perplexity’s practices, including the Chicago Tribune, News Corp, Encyclopedia Britannica, Merriam-Webster, Nikkei, Asahi Shimbun, and Reddit. Wired and Forbes have accused Perplexity of plagiarism and unethical scraping of content from websites that explicitly prohibit such activity, a claim corroborated by internet infrastructure provider Cloudflare.

Perplexity’s head of communications, Jesse Dwyer, has historically dismissed such lawsuits, suggesting they have never truly succeeded against new technologies. However, historical precedents show that publishers have, at times, won or significantly influenced legal battles over emerging technologies, leading to settlements and new licensing frameworks.

The current lawsuit against Perplexity follows a cease-and-desist letter sent by The Times over a year ago, with multiple attempts over 18 months to negotiate a content agreement. The Times’ ongoing suit against OpenAI and Microsoft involves claims that their AI systems were trained using millions of articles without compensation. OpenAI has argued “fair use” and accused The Times of manipulation. Meanwhile, a parallel lawsuit against OpenAI competitor Anthropic, concerning the use of pirated books for training, saw a court rule against using pirated content, resulting in a $1.5 billion settlement for Anthropic. This ruling could set a significant precedent for fair use in AI training.

Notably, The New York Times is open to partnerships with AI firms that provide fair compensation, having previously struck a multi-year deal with Amazon to license its content for AI model training. Similarly, other publishers such as Associated Press, Axel Springer, Vox Media, and The Atlantic have entered into licensing agreements with AI companies.