Significant Roadblocks for Plaintiffs in Generative Artificial Intelligence Lawsuit: California Judge Dismisses Most Claims Against AI Developers in Andersen v. Stability AI

On October 30, 2023, U.S. District Judge William Orrick of the Northern District of California issued an Order[ii] largely dismissing without prejudice the claims brought by artists Sarah Andersen, Kelly McKernan and Karla Ortiz in a proposed class action lawsuit against artificial intelligence (“AI”) companies Stability AI, Inc., Stability AI Ltd. (together, “Stability AI”), DeviantArt, Inc. (“DeviantArt”) and Midjourney, Inc. (“Midjourney”). 

Andersen is the first of many cases brought by high-profile artists, programmers and authors (including John Grisham, Sarah Silverman and Michael Chabon) seeking to challenge the legality of using copyrighted material for training AI models.

Background

Stability AI, DeviantArt and Midjourney all provide tools and services that use AI to generate new images based on text prompts entered by users.  The Complaint alleges that the Large-scale Artificial Intelligence Open Network (“LAION”), a German non-profit which makes open-sourced datasets, scraped over five billion images from the internet (including works by the plaintiffs) at the behest of Stability AI.[iii]  The Complaint further alleges that (i) Stability AI used this LAION dataset for the purpose of training its Stable Diffusion software product, (ii) DeviantArt and Midjourney incorporated Stable Diffusion into their products and (iii) Midjourney’s AI product was also “trained on a subset of the images used to train Stable Diffusion” (suggesting that Midjourney conducted its own unspecified training).[iv]  Claims brought against the defendants included direct copyright infringement, vicarious copyright infringement, violations of the Digital Millennium Copyright Act, violations of California’s right of publicity statute and common law rights of publicity and unfair competition under California state law.  All three defendants filed separate motions to dismiss, and DeviantArt filed a separate anti-SLAPP motion to strike the right of publicity claims, in which both Midjourney and Stability AI joined. 

In the first instance, Judge Orrick dismissed the copyright claims by Ortiz and McKernan with prejudice for lack of a valid copyright registration (a prerequisite to suit), and further narrowed Andersen’s copyright claims to only those images that had been registered with the Copyright Office at the time the suit was filed.[v]  The court also dismissed with prejudice plaintiffs’ claims for unfair competition to the extent based on alleged copyright infringement, which claims are preempted by the U.S. Copyright Act.  Judge Orrick granted leave to amend as to all other claims, but warned “I will not be as generous with leave to amend on the next, expected rounds of motions to dismiss and I will expect a greater level of specificity as to each claim alleged and the conduct of each defendant in support of each claim.”  Below are some key takeaways from the Order.

Key Takeaways

• Plaintiffs must prove actual unauthorized reproduction; mere usage of or reliance on an already trained model may not suffice for direct copyright infringement based on AI training.  The only direct infringement claim to survive defendants’ motions to dismiss was the direct copyright infringement claim against Stability AI based on its alleged copying and use of copyrighted images to train Stable Diffusion.  The direct infringement claims against DeviantArt and Midjourney were dismissed for failure to allege specific facts showing that they had, themselves, reproduced copyrighted images in training their models. Allegedly building platforms on Stability AI’s model was not sufficient to plead direct infringement.[vi] 

• Plaintiffs’ claims that AI outputs are  “derivative works”  failed in part for lack of substantial similarity to third-party copyrighted content.  Judge Orrick rejected plaintiffs’ argument that every output image from these generative AI tools must necessarily constitute a derivative work of the input data, given the implausibility that all training data is actually copyrighted or that all output actually relied on copyrighted training data.[vii]  Even so limited, Judge Orrick noted that plaintiffs must still show substantial similarity in protected expression between the original copyrighted work and specific output images, which may prove difficult given plaintiffs’ allegation that no outputs were “likely to be a close match for any specific image in the training data.”[viii] 

• Skepticism that the AI model itself could be a “derivative work.”   Judge Orrick expressed confusion regarding plaintiffs’ arguments that Stable Diffusion itself was a derivative work because it allegedly stored “compressed copies” of the copyrighted images it was trained on, and instructed plaintiffs to (i) clarify their theory as to how Stable Diffusion operates with respect to the training images, (ii) define “compressed copies” and (iii) allege plausible facts in support.[ix]  In their motions, defendants argued that their models—comprised of data and algorithms—could not plausibly be described as substantially similar in protected expression to any alleged copyrighted work on which they were trained, as necessary to assert a claim for violation of the derivative works right.  Judge Orrick deferred that determination until after plaintiffs amend.

• General claims regarding vicarious infringement or violation of DMCA and right of publicity do not suffice.  The court dismissed all claims for vicarious copyright infringement and violation of the DMCA and right of publicity with leave to amend. The vicarious infringement claims against DeviantArt and Midjourney were dismissed for failure to adequately plead any act of direct infringement.[x]  As to Stability AI, Judge Orrick reiterated that the claim required further clarity as to how Stable Diffusion supposedly stored “compressed copies” and made them available to alleged direct infringers (other platforms or users).  The court rejected any vicarious liability premised on the theory that all output is necessarily infringing, noting plaintiffs’ own allegations that no output is likely to be “a close match for any specific image in the training data.”[xi]  Similarly, Judge Orrick made clear that plaintiffs must allege plausible facts regarding what type of copyright management information was allegedly altered or removed from their works in violation of the DMCA, and which defendants allegedly performed such removing or altering.  And Judge Orrick rejected the adequacy of the right of publicity claims, noting that the Complaint failed to allege “any facts specific to the three named plaintiffs to plausibly allege that any defendant has used a named plaintiff’s name to advertise, sell, or solicit purchase of DreamStudio, DreamUp or the Midjourney product.”  The court did not resolve the anti-SLAPP motion directed to the publicity claims, but invited defendants to make the same challenge in response to any amended complaint.

This article was re-published by The Columbia Law School Blue Sky Blog as “Cleary Gottlieb Discusses Roadblocks for Plaintiffs in Generative Artificial Intelligence Lawsuit.” 

[i] Angela Dunning is a partner based in Cleary Gottlieb’s Bay Area office and Lindsay Harris is an associate based in Cleary Gottlieb’s New York office.  In addition to representing Midjourney in the matter discussed in this alert memorandum, Angela Dunning represents other leading AI developers in unrelated matters.

[ii] Order on Motions to Dismiss and Strike (the “Order”), Andersen v. Stability AI Ltd. et al, 3:23-cv-00201 (N.D. Cal. Oct. 30, 2023).

[iii] Complaint at 57, 101, 104, Andersen, 3:23-cv-00201.

[iv] Id. at *34-35, 57, 62, 101, 104, 115, 134, 148-49.

[v] Order at *5.

[vi] Id. at *8-14.

[vii] Id. at *10-12.

[viii] Orderat *11-12; Complaint at *93.

[ix] Order at *9.

[x] Id.

[xi] Id. at *15-16.