AI Art Lawsuit: Getty v. Stability, Rulings, and Fair Use

The AI art lawsuit landscape centers on a class action filed by visual artists against Stability AI, Midjourney, DeviantArt, and Runway AI, alongside a parallel suit brought by Getty Images. The artists’ case, Andersen v. Stability AI, is on track for a jury trial reported as scheduled for September 8, 2026 in the U.S. District Court for the Northern District of California, and it will be the first time a jury hears evidence on whether AI image generators infringe copyright by training on billions of scraped images.¹

Who Is Suing and What They Claim

On January 13, 2023, three artists — cartoonist Sarah Andersen, watercolor painter Kelly McKernan, and illustrator Karla Ortiz — filed a class-action complaint against Stability AI, Midjourney, and DeviantArt.² An amended complaint filed in November 2023 added Runway AI as a fourth defendant and expanded the plaintiff group to ten named artists. The plaintiffs are represented by Joseph Saveri, Matthew Butterick, and co-counsel Lockridge Grindal Nauen.³

The core allegation is that the defendants built their image generators by downloading and copying copyrighted artwork from LAION-5B, a dataset of roughly 5.85 billion image-text pairs scraped from the open web, without consent, credit, or compensation.⁴ The complaint calls the resulting tools “21st-century collage tools” that remix training data to produce new images on demand, often imitating specific artists’ styles.⁵

Two legal theories drive the copyright claims. The “model theory” argues the AI models are themselves infringing works because they contain compressed or mathematical representations of copyrighted images. The “distribution theory” argues that releasing those models is the equivalent of distributing the copyrighted works embedded inside.⁶⁵⁴

How the AI Companies Are Defending

The defendants argue that training does not involve copying or memorizing images, and that a finished model stores statistical relationships — weights and biases describing visual features — rather than literal copies of any artwork. They contend the training use is transformative and protected by fair use, and that generated outputs are not “substantially similar” to any specific copyrighted image, which is a requirement for infringement.⁴ Defenders of the AI companies have also emphasized that artistic style itself is not copyrightable.

What the Court Has Ruled So Far

The case, before Judge William Orrick, has been through two rounds of motions to dismiss.

On October 30, 2023, Judge Orrick dismissed most of the original claims but let Sarah Andersen’s direct copyright infringement claim against Stability AI proceed, giving the plaintiffs leave to amend.⁷ Claims against DeviantArt were largely gutted at this stage; the court found the plaintiffs had not shown DeviantArt played an affirmative role in scraping copyrighted works or training the Stable Diffusion model.⁸

The plaintiffs amended, and the defendants moved to dismiss again. On August 12, 2024, Judge Orrick issued the most consequential order in the case.⁹⁶¹⁰

Several claims were thrown out permanently. The DMCA claims under Sections 1202(a) and 1202(b) were dismissed with prejudice.⁹ Unjust enrichment claims were dismissed as preempted by the Copyright Act, and breach-of-contract and other peripheral theories were cut. The court did not decide the fair use question. That defense is being reserved for trial.¹⁰

Where the Case Stands Now

After the August 2024 ruling, the case moved into discovery. A third amended complaint was filed on February 27, 2026, and the defendants answered on March 13, 2026.⁷ The docket shows continued activity through at least May 2026, including discovery disputes and motions to compel.¹¹ Trial is reported as scheduled for September 8, 2026.¹

The Fair Use Question Hanging Over Everything

Whether training an AI model on copyrighted works is fair use is the legal question that will decide most of these cases, and courts have begun to answer it for other kinds of AI without touching image generators directly.

On June 23, 2025, Judge William Alsup of the Northern District of California ruled in Bartz v. Anthropic that using copyrighted books to train a large language model is “exceedingly transformative” and constitutes fair use. He rejected the argument that an emerging licensing market for AI training data entitled copyright holders to block the practice. The ruling came with a major limit: Anthropic’s use of pirated copies of books to build its training library was not fair use, which the judge called “inherently, irredeemably infringing.”¹² Judge Alsup also signaled that the analysis might come out differently for image generators, noting the plaintiffs in his case never alleged the AI outputs were infringing, and referencing “the Midjourney case” as a different scenario.¹³

Earlier, in February 2025, a Delaware federal court ruled in Thomson Reuters v. Ross Intelligence — the first case to reach a fair use determination on AI training — that copying legal headnotes to train an AI legal research tool was not fair use, because Ross’s product served the same purpose as the original and an “obvious” potential licensing market existed.¹⁴

These two rulings point in different directions, and neither involved image generators. Judge Orrick in Andersen has noted that “decisions in past cases—even other AI cases—are likely not to be influential” across different AI modalities.⁶ How fair use plays out for image generation is genuinely open.

Getty Images v. Stability AI

Getty Images has pursued Stability AI on two continents. In the UK, the case went to trial over several weeks in June 2025. On November 4, 2025, the High Court found that Stable Diffusion had committed trademark infringement by generating images containing Getty’s watermarks, though the court called these instances “limited” and “historic.” The court confirmed that Getty’s copyrighted works were used to train the model.¹⁵ Getty had abandoned its primary copyright training claim in the UK because it could not show the training took place within UK jurisdiction, and the court rejected the secondary copyright theory, holding that AI model weights are statistical parameters rather than stored copies of photographs.¹⁶ Getty has said it will carry the UK findings of fact into its US lawsuit.

The US case, filed in August 2025 in the Northern District of California, is early. A motion to dismiss is pending, mediation has been ordered for completion by October 2026, and a jury trial is scheduled for early 2028.¹⁷

Other AI Copyright Cases in Motion

The image-generator cases are one slice of a broader docket that covers text, code, music, and news content. Rulings in the non-image cases will influence the legal environment even though they do not decide the visual-art question directly.

In re: OpenAI Copyright Infringement Litigation, a multidistrict litigation in the Southern District of New York consolidating 12 cases against OpenAI and Microsoft over training on copyrighted text.¹⁸
Concord Music Group v. Anthropic, alleging that Anthropic’s Claude chatbot was trained on copyrighted song lyrics without authorization.¹⁸
Doe v. GitHub, the Copilot litigation, alleging GitHub, Microsoft, and OpenAI used open-source code to train Codex and Copilot in violation of licensing terms and the DMCA. It is currently stayed for an interlocutory appeal.¹⁸
Dow Jones v. Perplexity AI, brought by news publishers over the AI search startup’s use of their content in its retrieval-augmented generation system.¹⁸

Copyright Office and Congress

The U.S. Copyright Office released a pre-publication version of its Part 3 report on May 9, 2025 addressing AI training. The report declined to call AI training categorically fair use. It found that when a model is trained to produce content sharing the purpose of the underlying works, such as entertainment, the use is “at best, modestly transformative.” It flagged market dilution under the fourth fair use factor and rejected the analogy between AI training and human learning, citing the differences in scale, speed, and reproduction. The Office recommended letting voluntary licensing markets develop and identified collective licensing as a potentially “workable” solution in industries such as music and stock photography.¹⁹

Congress has not passed legislation on AI training and copyright. The TRAIN Act (S. 2455), introduced in July 2025 by Senator Peter Welch with bipartisan co-sponsors, would create an administrative subpoena process letting copyright owners identify which of their works were used in AI training.²⁰ A House version was introduced in January 2026.²¹ The bill remains early in the legislative process, with no committee votes or markup on record, and analysts have flagged the risk of a presidential veto given the current administration’s emphasis on AI deregulation.²²

What to Watch Next

The September 2026 Andersen trial, if it holds, will put two questions to a jury for the first time: whether image-generation models contain infringing copies of copyrighted artwork, and whether the companies that built them can win on fair use. The trial will also test whether the model theory holds up under full evidentiary scrutiny after surviving two rounds of motions to dismiss. Getty’s US suit is behind it, with a trial set for early 2028, and the multidistrict OpenAI litigation is on its own track. For now, the rules governing AI’s use of creative work are unsettled, and early rulings point in competing directions depending on the type of content, the type of AI, and the specific facts of each case.

1
NYU JIPEL. Andersen v. Stability AI: The Landmark Case Unpacking the Copyright Risks of AI Image Generators
2
CourtListener. Andersen v. Stability AI Ltd., Case No. 3:23-cv-00201
3
Image Generator Litigation. Image Generator Litigation
4
It’s Art Law. Artificial Intelligence and Artists’ Intellectual Property
5
IPWatchdog. Andersen et al v. Stability AI Class Action Complaint
6
Copyright Alliance. Andersen v. Stability AI Copyright Case
7
Meshi P Law. Andersen, et al. v. Stability AI Ltd., et al.
8
The Fashion Law. Court Refuses to Dismiss Artists’ Copyright Claim Against Stability AI
9
Justia. Andersen et al v. Stability AI Ltd. et al, Order on Motions to Dismiss
10
The Art Newspaper. US Artists Score Victory in Landmark AI Copyright Case
11
CourtListener. Andersen v. Stability AI Ltd., Docket Page 3
12
Goodwin. District Court Issues AI Fair Use Decision
13
AFS Law. Landmark Ruling: AI Copyright Fair Use vs. Infringement, Bartz v. Anthropic
14
Copyright Alliance. AI Training Not Fair Use
15
Getty Images. Getty Images Issues Statement on Ruling in Stability AI UK Litigation
16
Mayer Brown. Getty Images v. Stability AI: What the High Court’s Decision Means
17
CourtListener. Getty Images (US), Inc. v. Stability AI, Ltd., Case No. 3:25-cv-06891
18
Baker Hostetler. Case Tracker: Artificial Intelligence, Copyrights, and Class Actions
19
Copyright Alliance. Copyright Office’s AI Report Takeaways
20
Copyright Alliance. Copyright Congress 2025 Year in Review
21
U.S. Copyright Office. Copyright Office Legislative Developments
22
Berkeley Technology Law Journal. The TRAIN Act: Forcing Transparency in AI Training Data