The New York Times has filed a sweeping copyright lawsuit against Perplexity AI, accusing the startup of “illegal” copying of millions of its articles and other protected works in order to build and power its products. The complaint, lodged in a U.S. federal court, alleges that Perplexity’s systems systematically scraped, stored and reproduced Times journalism without permission, striking at the heart of how news organizations monetize their reporting.
The case marks a sharp escalation in the confrontation between major media outlets and fast-growing AI firms, arriving after a separate suit by the Chicago Tribune over similar practices and signaling that publishers are increasingly prepared to test AI training and output in court rather than rely on voluntary licensing deals.
The Filing Details
The New York Times brought its case against Perplexity AI in a U.S. federal court, with the filing on 5 December 2025 framed as a pivotal moment in the legal fight over how generative AI systems ingest and reuse news content. According to reporting on the complaint, the lawsuit asserts that Perplexity infringed copyright works owned by the Times by copying and processing vast quantities of its journalism to train and run the company’s conversational search tools and related services, without any license or negotiated agreement in place. By anchoring the filing to a specific federal venue and date, the Times is treating this as a test case that could shape how courts view large-scale scraping of subscription and ad-supported news sites for AI purposes, a question that has become central to the business models of both publishers and AI startups.
Coverage of the complaint describes how the Times accuses Perplexity of “illegal” copying that goes beyond incidental indexing or brief quotation, alleging that the startup’s technology ingested and reused millions of full articles and other protected works. In that telling, the lawsuit argues that Perplexity’s systems did not simply learn from high-level patterns in the data but instead stored and reproduced Times content in ways that substitute for reading the original stories, undermining the value of the Times’ paywalled reporting and its investments in investigative and international coverage. By casting the case as a direct challenge to unauthorized reproduction at scale, the Times is signaling to other AI firms that unlicensed scraping of premium journalism could now carry significant legal and financial risk.
Allegations of Copyright Infringement
At the core of the complaint are detailed allegations that Perplexity AI copied New York Times articles without consent and then used that material to generate outputs that closely track the original reporting. According to a summary of the lawsuit, the Times contends that Perplexity’s systems ingested millions of its articles and other works, including content that sits behind a paywall, and then reproduced that material in user-facing answers that compete directly with the Times’ own digital products. The filing characterizes this conduct as “illegal” copying that violates the Times’ exclusive rights to reproduce and distribute its journalism, arguing that Perplexity’s approach effectively treats the publisher’s archive as a free training set and content feed rather than a licensed resource.
Reporting on the case notes that the Times points to specific examples where Perplexity’s responses allegedly mirrored Times content, including passages that track the structure and language of original stories so closely that they function as substitutes rather than summaries. In those instances, the complaint asserts that Perplexity’s outputs did not merely paraphrase or reference Times reporting but instead reproduced protected expression in a way that would allow users to bypass the Times’ website and apps entirely. If a court agrees that such outputs amount to direct infringement rather than transformative use, the ruling could set a precedent that forces AI companies to either secure licenses for large news archives or redesign their systems to avoid regurgitating source material.
Prior Lawsuits Against Perplexity AI
The New York Times lawsuit lands on top of existing legal pressure from other publishers, most notably a case brought by the Chicago Tribune that also targets Perplexity’s handling of news content. According to a report on that earlier action, the Chicago Tribune accused Perplexity of copying and reusing its stories without authorization in a way that paralleled the Times’ new claims, arguing that the startup’s tools surfaced Tribune reporting to users while bypassing the paper’s own digital platforms. By the time the Times filed its complaint, Perplexity was already defending itself against allegations that its core technology depends on unlicensed scraping of journalism, a pattern that could weigh on investor confidence and partnership talks as the company seeks to grow.
Coverage of the Tribune case highlights how both lawsuits frame Perplexity’s conduct as “illegal” copying that threatens the economic foundations of local and national newsrooms. In each instance, the publishers argue that Perplexity’s products free ride on the costs of original reporting, editing and distribution, while diverting audience attention and potential subscription revenue away from the outlets that produced the underlying work. The emergence of multiple, similarly framed lawsuits in the wake of 2025 developments suggests that major news organizations are beginning to coordinate their legal strategies, using parallel cases to test common arguments about scraping, training data and AI-generated outputs across different jurisdictions and fact patterns.
Broader Industry Ramifications
Legal analysts following the case say the New York Times lawsuit has the potential to influence how courts and regulators treat other AI startups that rely on large-scale scraping of news sites and other content repositories. According to a detailed account of the filing, the Times is not only seeking damages for past copying but also asking the court to clarify the boundaries of permissible use of copyrighted works in AI training and generation, particularly when the works are paywalled or otherwise restricted. If the court finds that Perplexity’s ingestion and reproduction of Times content infringed copyright, similar claims could be brought against a wide range of AI firms that have built models on top of web-scale datasets assembled without explicit permission from rights holders.
Reporting on the complaint underscores that the Times is focusing attention on Perplexity’s business model, which allegedly depends on access to millions of articles and other works that the publisher views as proprietary assets rather than raw material for open-ended machine learning. By highlighting the scale and specificity of the copying, the Times is effectively challenging the assumption that scraping publicly accessible web pages for AI training is legally and commercially acceptable, especially when the resulting products compete with the original sources for readers and advertisers. That argument resonates with broader tensions between publishers and technology companies in the AI era, as news organizations seek to ensure that their content is not quietly transformed into a commodity input for tools that could erode their audiences and bargaining power.
How the New York Times Frames the Stakes
In its public framing of the lawsuit, the New York Times presents the case as a defense of both its own intellectual property and the broader ecosystem of professional journalism. One detailed report on the complaint explains that the Times accuses Perplexity of “illegal” copying of millions of articles, including investigative pieces and international coverage that required significant investment of time and resources, and argues that unlicensed AI reuse of that work undermines the incentives to fund similar reporting in the future. By emphasizing the volume and value of the allegedly copied material, the Times is positioning the dispute not as a narrow technical disagreement over fair use, but as a fundamental clash over who controls and profits from high quality news content in a market increasingly shaped by AI-driven discovery tools.
Another account of the filing notes that the Times links its claims against Perplexity to a wider concern that AI systems trained on unlicensed news content can distort or misrepresent original reporting while still drawing on the authority of established outlets. The complaint points to instances where Perplexity’s outputs, while closely tracking Times articles, may omit context, nuance or corrections that are present in the original stories, raising the risk that users receive incomplete or misleading information that they nonetheless associate with the Times’ brand. By tying copyright enforcement to questions of accuracy and trust, the Times is arguing that publishers have a stake not only in how their content is monetized, but also in how AI tools present and adapt their work to audiences who may never click through to the source.
Implications for AI Governance and Future Deals
The outcome of the New York Times lawsuit is likely to shape how AI companies approach licensing negotiations with publishers, as well as how policymakers think about updating copyright rules for the age of generative models. Detailed coverage of the case suggests that the Times is seeking remedies that could include monetary damages and injunctive relief limiting Perplexity’s use of its content, outcomes that would send a strong signal to other AI firms that unlicensed scraping of premium news archives carries substantial downside risk. Faced with that possibility, startups and larger platforms alike may feel increased pressure to strike formal licensing agreements with major news organizations, similar to the deals some publishers have already reached with search engines and social networks for display of headlines and snippets.
At the same time, the lawsuit is feeding into a broader policy debate over whether existing copyright doctrines, including fair use, adequately address the realities of AI training and generation. If courts conclude that large-scale ingestion of copyrighted works for model training is permissible but that regurgitation of protected text in outputs is not, AI developers may need to invest in new technical safeguards to prevent their systems from reproducing source material too closely. Conversely, if judges take a more restrictive view of training data itself, the industry could face a wave of claims from rights holders across media, entertainment and publishing, accelerating calls for legislative reforms or collective licensing frameworks that balance innovation with compensation for creators.
Perplexity AI Under Intensifying Scrutiny
Although the current reporting focuses primarily on the New York Times’ allegations, the lawsuit adds to a growing narrative that Perplexity AI is under intensifying scrutiny from traditional media companies. One summary of the dispute notes that Perplexity was already being sued by the Chicago Tribune for similar content copying issues, and that the Times case compounds the legal and reputational challenges facing a startup that has marketed itself as a next generation answer engine built on top of the open web. With multiple prominent publishers now questioning the legality of its data practices, Perplexity must not only defend itself in court but also reassure potential partners and users that its technology respects intellectual property and supports, rather than undermines, the news ecosystem.
The convergence of these lawsuits also raises questions about how investors and larger technology companies will evaluate partnerships and acquisitions involving AI firms that rely heavily on scraped content. If courts side with the Times and the Tribune, the valuation of companies built on similar data strategies could be affected, and due diligence processes may place greater emphasis on documented licenses and content sourcing practices. For news organizations, the early wave of litigation against Perplexity serves as both a warning and a template, illustrating how publishers might use the courts to push back against AI tools that they believe are eroding their business models while capitalizing on their work.
