By Professor Robert J. Romano, JD, LLM, St. John’s University
On June 11, 2025, Disney Enterprises, Inc, together with several of its affiliates (Marvel Characters, Inc., Lucasfilm, LTD, and Universal City Studios, to name a few), threw Thor’s Hammer down and filed suit in the U.S. District Court, Central District of California, against one of the leading artificial intelligence (AI) image generators, Midjourney, Inc. The two count, one-hundred-and-forty-three page complaint, which includes numerous images of Star War’s Darth Vader and Yoda, various Simpsons characters, and Marvel’s Spiderman, Hulk, Black Panther, and Iron Man, amongst others, claims direct and secondary copyright infringement in that Midjourney went Rogue and reproduced, publicly displayed and distributed Disney’s copyrighted works in violation of Section 106 of the Copyright Act.[1]
For those of us over the age of thirty who don’t know who Midjourney is or what it does, it is a company that has established an AI tool which allows a user to create realistic images from written text descriptions, or what is referred to as “prompts”. That user can basically describe any image he or she wants, and Midjourney’s AI system generates a graphic in an assortment of styles, such as realistic, painterly, or cinematic. The AI service runs as a “bot” on what is known as the “Discord” platform that allows for users to generate images, create variations, and even upscale them to higher resolutions. Disney alleges that Midjourney Trained its Dragon of an AI system to create images of its famous characters, developing what is referred to in its complaint as a “virtual vending machine” and “bottomless pit of plagiarism” that spewed “endless unauthorized copies of Disney’s and Universal’s copyrighted works.”[2]
In order to Guard its Galaxy of copyrighted images and prove direct infringement, Disney and its Minions must get to the Groot of the issue and establish that it owns the copyrighted work and then prove that Midjourney copied the protected elements of that work without Disney’s authorization (Or as Yoda would say – “Used my image without permission you did.”). Copying can be established through either direct or indirect evidence, which requires proof that the defendant had “access” to the copyrighted work and that the two works are “substantially similar”. Interestingly, there are no provisions within the Copyright Act regarding secondary infringement, however, but judges have Used the Force and established that such a cause of action can be asserted against a party that does not directly infringe on a copyright owner’s rights but does partake in the facilitating of an infringement by others. The elements of secondary infringement are similar to direct infringement but encompass two principal forms: contributory infringement and vicarious liability. Contributory infringement can be found when a party, with knowledge of the infringing activity, induces, causes or materially contributes to the infringing material, while as for vicarious liability, a party must have both the right and ability to supervise or control the infringing action of the direct infringer, while also receiving a direct financial benefit from the infringing activity. Actual knowledge of an infringement is immaterial when determining vicarious liability. It should be noted, however, that a direct infringement claim must be established first before either a contributory or vicarious claim can move forward.
The Thing is, litigation is like an onion, it has layers, and Midjourney, in response to Disney’s one-hundred- and forty-three-page complaint, went to Infinity and Beyond and filed its own forty-three-page response that takes issue with Disney for trying to have it both ways, seeking to profit—through their use of Midjourney and other generative AI tools—on the one hand, while on the other hand accusing Midjourney of wrongdoing for the same.[3] More importantly, however, Midjourney argues that Disney is wrong on the merits of the case since copyright law does not confer absolute control over the use of copyrighted works. Midjourney argues that “The limited monopoly granted by copyright must give way to fair use, which safeguards countervailing public interests in the free flow of ideas and information,” and that “Training a generative AI model to understand concepts by extracting statistical information embedded in copyrighted works is a quintessentially transformative fair use – a determination resoundingly supported by courts that have considered the issue.”[4]
In the end, as with all court cases, nothing moves like Lightning McQueen, where things get bogged down and Frozen within the various nuances of litigation. Therefore, since this case can possibly test the boundaries of copyright law for generative AI and could set a precedent for how the AI industry is regulated, hopefully, both parties can come to the table and come up with a reasonable solution that can benefit all of those involved. Full disclosure – without the help of either Midjourney or Disney, this article was AI generated. (Or, as Linda McMahon, the Secretary of the Department of Education, would say – “A1 generated”).[5]
[1] Case 2:25-cv-05275 Document 1 Filed 06/11/25.
[2] Id.
[3] Id at page 3.
[4] Id. See Kadrey v. Meta Platforms, Inc., 2025 WL 1752484, at *12 (N.D. Cal. June 25, 2025) (describing training as “highly transformative”); Bartz v. Anthropic PBC, 2025 WL 1741691, at *7 (N.D. Cal. June 23, 2025).
[5]https://www.google.com/search?sca_esv=1f08cea1075aad4d&udm=7&q=linda+mcmahon+a1+generated&sa=X&ved=2ahUKEwjX1av33rWPAxUEKVkFHRHIBz0Q8ccDKAR6BAgWEAY&biw=1424&bih=727&dpr=2#fpstate=ive&vld=cid:bc5a5531,vid:6QL0c5BbCR4,st:0
