AI Copyright Lawsuit Developments in 2025: A Year in Review
The rapid development of generative artificial intelligence (AI) models over the past few years has given rise to now over 70 infringement lawsuits by copyright owners against AI companies. While the cases are in different stages of litigation, there were several big developments—including orders on summary judgment, high-profile settlements, and dozens of new cases filed—in 2025 that give us some insights into how courts and litigants are navigating generative AI copyright infringement issues. This blog will focus on the most significant takeaways from 2025 and look ahead to what is sure to be a pivotal year for AI-related copyright lawsuits in 2026.
Two Major Decisions on Fair Use and One Big Settlement
There is little doubt that the biggest lawsuit development of 2025 was the $1.5 billion settlement in theBartz v. Anthropic case—a case in which Anthropic faced a potentially massive statutory damages penalty for downloading millions of pirated copies of works it used for training. But it’s important to understand how the parties got there and how a parallel decision in Kadrey v. Meta compares.
Bartz v. Anthropic
Back on June 23, a district court in the Northern District of California issued an order on summary judgment finding that Anthropic’s generative large language model (LLM) training was “exceedingly” transformative and thus qualified as fair use. At first blush, many commentors and journalists chalked it up as a big win for generative AI companies, but in addition to the decision’s many flaws, it became clear that that was not the case as Anthropic was still on the hook for potential damages in the billions from its downloading of pirated works. As a likely result of such potentially mammoth liability, in September the parties announced a $1.5 billion settlement had been reached, requiring Anthropic to pay approximately $3,000 for each of the 482,460 books it downloaded from the pirate libraries Library Genesis, and Pirate Library Mirror.
It was a landmark moment that sent shockwaves through the AI copyright litigation world, where many AI developer defendants have been alleged or admitted to using training material from the same piracy-laden datasets. Copyright Alliance CEO Keith Kupferschmid stated at the time that “[w]hile the settlement amount is very significant and represents a clear victory for the publishers and authors in the class, it also proves what we have been saying all along—that AI companies can afford to compensate copyright owners for their works without it undermining their ability to continue to innovate and compete.”
Kadrey v. Meta
Just two days after the Bartz summary judgment order, on June 25, another district court in the Northern District of California issued an order on summary judgment in a case brought by book authors against Meta, finding that the use of the plaintiffs’ books for training Meta’s LLM was “highly transformative” and qualified as fair use. While the holding appeared to be detrimental to the plaintiffs in the case, the overall impact was questionable because the court made abundantly clear that its decision was very narrow due to it being based on a lack of evidence presented by counsel.
What will likely prove to be the most important aspect of the decision was the court’s thoughtful and lengthy discussion of the indirect substitutional impacts that could harm the copyright owners’ actual and potential markets. Unlike Bartz, the Kadrey court understood and appreciated that the advent of generative AI technology and the scale of its impact on copyright’s incentives to create and distribute new copyrighted works for the public to enjoy requires a more thoughtful approach that is more consistent with the spirit and purpose of copyright and the fair use defense.
While the summary judgment order was a setback for the plaintiffs, the case is proceeding on the question of whether Meta simultaneously uploaded copyrighted works while using BitTorrent technology to download them—a process known as seeding. If the court finds that Meta distributed massive amounts of pirated copyrighted works, it could be held liable for staggering damages (similar to Anthropic) and perhaps another big settlement will be reached in 2026. But what may be the most important aspect of the case is the roadmap to future copyright-owner plaintiffs on how to prevail in their AI infringement cases that was explicitly provided by the judge in the case.
Other Settlements Signal a Trend
In the months following the Bartz settlement, numerous other agreements were reached in high-profile cases in the AI music space, some of which included partnerships and licensing deals between AI developers and copyright owners.
Record Label Cases
On October 29, Universal Music Group (UMG) announced that it had settled its AI copyright infringement lawsuit against Udio. In addition to the compensatory legal settlement, the parties entered into license agreements for UMG’s recorded music and publishing catalogues. According to the announcement, the two parties will work on launching a new subscription service in 2026 for a generative AI technology trained on fully authorized and licensed music. The license is reportedly structured on artists’ opt-in, giving copyright owners and creators the control over their works guaranteed by the law, rather than an unworkable “opt-out” option that many AI companies have promoted.
Not long after the UMG settlement, on November 19, Warner Music Group (WMG)—which was also party to the Udio lawsuit—announced that it had also settled. Like UMG, the settlement agreement includes a licensing deal where WMG artists can opt-in to license their works to develop Udio’s new subscription service, which will be launched in 2026. It’s important to note that while UMG and WMG reached an agreement with Udio, the other record label plaintiff, Sony, has not, and the case is ongoing.
In November, Warner Music Group (WMG) and Suno, another leading music AI company, announced that the parties settled a lawsuit filed by a group of record labels against Suno over the unlicensed use of music from WMG’s repertoire to train Suno’s AI model. According to the announcement, Suno will launch an entirely new model in 2026 that consists of “more advanced and licensed models” while current models will be phased out. The announcement also explains that “[a]rtists and songwriters will have full control over whether and how their names, images, likenesses, voices, and compositions are used in new AI-generated music.”
These agreements, along with other confidential settlements in Vacker v. ElevenLabs and Planner 5D v. Facebook that were reached in 2025, demonstrate that AI companies and copyright owners are more willing than ever to come to the table and negotiate AI licensing deals. Indeed, it’s been reported that there have been ongoing settlement negotiations in the In re OpenAI litigation, which involves over a dozen centralized cases and could have a huge impact on LLM litigation. It’s safe to say that settlements and partnerships were the big trend in AI copyright litigation in 2025, and they’re likely to multiply in 2026.
Steady Stream of New Cases
The number of infringement cases filed against AI companies in 2025 more than doubled the total at the end of 2024, from around 30 to now over 70. While it would be too difficult to cover all the new cases in this blog, here are some of the highlights.
In Re OpenAI (aka the Multidistrict Litigation)
While not technically a newly filed case, it’s worth mentioning that 2025 saw the centralization of over a dozen cases brought by literary and news media copyright owners against OpenAI in the Southern District of New York. The cases are now overseen by Judge Sidney Stein for pretrial activity, which could include dispositive motions including summary judgment, as well as discovery disputes that have been common among the cases. The multidistrict litigation includes high-profile cases like New York Times v. OpenAI and Authors Guild v. OpenAI, which, while still technically separate (not consolidated), could be subject to orders in 2026 that impact all cases involving large language models. It should also be noted that there have been settlement negotiations reported that could resolve some of the In Re OpenAI cases in 2026.
Film Studios Cases
2025 saw major movie studios enter the litigation arena against AI companies for the first time. In June, Disney and Universal filed a complaint against Midjourney in the Central District of California for direct and secondary copyright infringement related to the company’s AI image generating service. The complaint alleges that Midjourney reproduces, publicly displays, and distributes copies and derivatives of characters from popular Marvel and Star Wars franchises, among others, and it provides visual evidence of dozens of instances of infringing output.
In September, Warner Bros. Entertainment filed a similar complaint against Midjourney over the unauthorized use of its copyrighted works to train Midjourney’s AI models, alleging that Midjourney actively knew about such outputs but failed to implement safeguards to prevent the generation of infringing outputs. On November 4, the Warner Bros. case was consolidated with Disney’s case as Disney Enterprises, Inc., et al.; Warner Bros. Entertainment, Inc., et al., v. Midjourney, Inc.
Also in September, Disney, Universal, and Warner Bros. filed the first AI copyright infringement lawsuit against a foreign company, the China-based Minimax. The complaint alleges Minimax used the movie studios’ copyrighted works to train the AI image and video generator, Hailuo AI, and that the model can be prompted to generate image and video outputs containing copyright-protected characters from beloved franchises including Star Wars, The Simpsons, Despicable Me, Shrek, Scooby Doo, and Looney Tunes. So far, the movie studio plaintiffs have faced obstacles attempting to serve the foreign based defendants, which will be an issue to watch in 2026.
RAG Cases Against Perplexity and Cohere
The past year also saw a number of new cases filed against Perplexity, an AI company that relies heavily on retrieval augmented generation (RAG) technology to provide real-time responses scraped from material on the internet. In September, Encyclopedia Britannica, Inc. and Merriam-Webster, Inc. filed a complaint against Perplexity over the unauthorized use of its publications and works for the AI company’s RAG services. The plaintiffs allege that their websites were crawled and scraped by PerplexityBot, their copyrighted articles were copied at the input stage in response to user queries, and their copyrights were infringed at the output stage where the RAG model generates output substantially similar to those inputs.
Then, in early December, Perplexity was hit with back-to-back lawsuits from the Chicago Tribune and the New York Times. Both complaints allege that the respective news media companies’ content was used without authorization to train Perplexity AI’s platforms including for RAG purposes. Given their similarity, the two cases are likely to be consolidated.
Another AI company that employs RAG technology, Cohere, was hit with a lawsuit from Advance in early 2025 for the unlicensed copying, use, and dissemination of Advance’s news and magazine articles to train Cohere’s suite of large language model AI systems called the “Command Family” of models. According to the complaint, Cohere’s AI models deliver outputs that include full verbatim copies, substantial excerpts, and substitutive summaries of the publishers’ works. The publishers also point out that the retrieval-augmented generation (RAG) feature in Command routinely returns verbatim copies of publishers’ copyright protected works in response to user queries.
These RAG-focused cases were supplemented by news publishers in Japan and Italy that brought similar suits against Perplexity in 2025. This year will likely see even more attention paid to AI models that largely rely on RAG technology, but given that last year also saw partnerships announced between Perplexity and copyright owners like Getty Images, there may be settlements on the horizon.
New Big-Name Defendants
A handful of big companies for the first time found themselves on the wrong end of AI infringement lawsuits in 2025. In September, a group of authors filed a complaint against Apple over the company’s unlicensed use of books to train its large language generative AI models, OpenELM. Then in October, another group of authors led by Tasha Alexander brought a class action lawsuit against Apple and senior executives at the company over the unlicensed use of copyrighted books to train its OpenELM. Both complaints allege that Apple trained its AI models using the RedPajama dataset, which contains the Books3 dataset—a massive online dataset known to contain pirated copies of books sourced from pirate e-repositories including Library Genesis, Z-Library, Sci-Hub, and Anna’s Archive.
In November, the same group of authors, led by Tasha Alexander, filed a class action lawsuit against Salesforce over the unlicensed use of their books to train Salesforce’s CodeGen and XGen series of large language models (LLMs). That lawsuit followed a class action lawsuit brought in October by another group of authors, led by E. Molly Tanzer and Jennifer Gilmore, against Salesforce for using their books to train Salesforce’s XGen series of large language models (LLMs). Both lawsuits allege that the training datasets used to develop the XGen LLM included Red Pajama and The Pile.
In December, author Elizabeth Lyon filed a class action lawsuit against Adobe over the unlicensed use of textual works to train Adobe’s small language model (SLM) SlimLM. Lyon alleges that Adobe used the pre-training dataset, SlimPajama that is derived from the RedPajama dataset containing copies of books sourced from pirate websites.
Lastly, in December, Ted Entertainment, which is a group of video creators, sued ByteDance, the parent company of TikTok, alleging the Chines social media giant unlawfully circumvented technological measures to access and scrape millions of copyrighted videos from YouTube in order to feed, train, improve, its generative artificial intelligence model, “MagicVideo.” Thus, rather than bring direct copyright infringement claims, the video creators sued for violations of the DMCA’s 1201(a) anticircumvention provisions.
Indie Musicians’ Lawsuits
While some record labels settled and entered into partnerships with AI music generators last year, groups of independent musicians banded together in 2025 to sue the same AI companies. In June, country music artist Anthony Justice (aka “Tony Justice”) filed a complaint against AI music generator Suno, and a similar lawsuit against Udio, on behalf of a group of musicians, alleging the companies used his sound recordings without authorization to train their model and that the models reproduce exact or near exact replicas of his songs. Both AI companies have since filed motions to dismiss, which will be the focus of the lawsuits in early 2026.
In October, another group of musicians, led by David Woulard, filed a class action lawsuits against both Udio and Suno over the unlicensed use of plaintiffs’ sound recordings and musical works to train the companies’ AI model. The plaintiffs claim that Udio and Suno scraped lyrics websites and also engaged in stream ripping songs from YouTube. In December, the same group of plaintiffs, in addition to a few other musicians, brought similar claims in a class action lawsuit against Chinese company, Kunlun Tech. These class action lawsuits brought by independent musicians, as well as other music cases, now include DMCA violation claims related to stream ripping that represent a new litigation strategy that will be something to keep an eye on in 2026.
Looking Ahead
While there may be some orders on various issues over the first few months of the new year, we’ll likely have to wait until well into 2026 to get the next decisions on fair use, which is the primary issue in almost all of these cases. According to this chart, the next few cases that will see decisions on fair use (unless settled or subject to rescheduling/extensions) are In re Google Generative AI, UMG v. Suno, Concord v. Anthropic, and In re Mosaic LLM Litigation—but even those decisions are not likely to start being issued until summer 2026, at the earliest. However, there are other developments to keep an eye on.
Thomson Reuters v Ross
This year will likely see a decision in a long-running copyright infringement case brought by Thomson-Reuters (TR) against Ross Intelligence, a competitor legal research service, for copyright infringement related to Ross’s scraping of copyrighted legal content from TR’s Westlaw platform to develop its own competing product based on machine learning. While the case doesn’t involve the same type of generative AI technology at the heart of many of the large language model or image generator lawsuits, it has raised important questions about copyrightability and fair use as applied to AI training.
In early 2025, the district court granted TR’s motion for summary judgment on direct copyright infringement and related defenses against Ross. The court found that TR’s Westlaw headnotes are copyrightable and rejected Ross’ fair use defense, explaining that the use was commercial and not transformative (under factor 1) and that Ross’ use harmed the potential market for AI training data (under factor 4). However, in April, the court granted Ross’s motion for interlocutory appeal to the Third Circuit on the issues of copyrightability and fair use.
In late September, a number of amicus briefs were filed in support of Ross, followed in November by many briefs in support of Thomson Reuters, including a Copyright Alliance brief. A detailed summary of the briefs in support of Thomson Reuters can be found here. The next step in the case is oral arguments, though a date has not yet been set.
Conclusion
The settlements of 2025 provide important precedent for the value of copyright claims in the AI context and demonstrate different pathways towards resolution. Additionally, the record label partnerships show that litigation can lead to collaborative arrangements rather than just monetary judgments. However, the uptick in new cases filed, which for the first time include major film industry plaintiffs, as well as companies that had avoided being sued until last year, show that there’s still a lot of disputes to be resolved in the years ahead. Even though the next major fair use decisions are not expected until this summer, 2026 is sure to be a pivotal year for AI and copyright issues in the courts.
If you aren’t already a member of the Copyright Alliance, you can join today by completing our Individual Creator Members membership form! Members gain access to monthly newsletters, educational webinars, and so much more — all for free!