The red carpet has officially rolled into federal court.
In a move that could reshape how copyright law applies to artificial intelligence, Disney and Universal, two giants of global entertainment, have sued the AI image-generation platform Midjourney. Filed on June 14th in Los Angeles, the lawsuit alleges that Midjourney enables users to generate strikingly realistic depictions of iconic characters like Yoda, Spider-Man, Elsa, and Shrek, without permission and to the detriment of the original creators.
Hollywood Enters the AI Copyright Arena
The complaint marks a turning point in the entertainment industry’s response to generative AI. Once a distant observer of the AI boom, Hollywood is now stepping into the legal arena, challenging the legitimacy of training methods and outputs that rely on copyrighted material.
Midjourney creates its images by processing user-submitted prompts using vast datasets scraped from across the internet. According to the studios, these datasets include copyrighted content. In many cases, the resulting images depict recognizable characters in familiar poses, costumes, or settings, raising the question of whether this constitutes inspiration or outright infringement.
Key Claims in the Disney and Universal Lawsuit
The lawsuit advances four key claims:
- Midjourney enables the creation of unlicensed, substantially similar reproductions of copyrighted characters.
- It failed to implement effective filters to block infringing prompts despite repeated takedown requests.
- It continued to host and facilitate infringing content, in violation of the DMCA.
- Its outputs undermine the market for licensed works, diluting the value of original content.
Anticipated Fair Use Defense from Midjourney
Midjourney has not yet responded publicly, but it is widely expected to raise a fair use defense, arguing that AI-generated outputs are “transformative.” Under U.S. copyright law, a work is considered transformative if it adds new expression, meaning, or purpose to the original. Courts have occasionally accepted this defense when a work repurposes or comments on the original in a new and meaningful way.
But courts have grown more cautious, especially when AI outputs closely resemble their source material. In Thomson Reuters v. Ross Intelligence, a federal court in Delaware recently held that using copyrighted legal headnotes to train an AI system was likely not fair use. This ruling, along with pending cases brought by Getty Images and The New York Times, suggests the judiciary may be drawing firmer boundaries around what AI systems can legally ingest and reproduce.
Why This AI Copyright Lawsuit Matters
The outcome of Disney’s lawsuit could carry major implications for the AI industry. A ruling in the studios’ favor may force platforms to obtain licenses for training data, impose stricter prompt filters, and proactively block infringing content. Conversely, a win for Midjourney could expand the scope of fair use, encouraging developers to continue training models on publicly available content, regardless of copyright status.
At the heart of the issue is the way these AI systems are built. Platforms like Midjourney rely on scraping enormous amounts of internet content—artwork, photos, and text—often without the consent of the creators. The challenge is structural: training large models effectively requires access to broad and diverse datasets. Without it, the models lose accuracy, context, and creative range. Yet the legal framework has not kept pace, leaving courts to determine whether this foundational practice is legally defensible.
Broader Impact on the AI and IP Landscape
Other lawsuits are surfacing similar tensions. In Getty Images v. Stability AI, Getty alleges that millions of its copyrighted photographs were scraped to train an image-generation model. In The New York Times v. OpenAI and Microsoft, the Times claims its articles were used without consent to train ChatGPT and other large language models. These cases—and now Disney v. Midjourney—all converge on a pivotal and unresolved legal question: can AI companies use copyrighted material to develop and operate their tools without the rights holder’s permission?
The answer may redefine the boundaries of copyright enforcement in the AI era. As Midjourney unfolds, stakeholders across the creative and tech industries should closely monitor how courts interpret fair use, transformative expression, and market harm in the context of machine-generated content.
A win for Disney and Universal could compel AI companies to fundamentally alter how they source, filter, and license data. A win for Midjourney could validate a more expansive reading of fair use, potentially accelerating AI development, but also deepening concerns around authorship, originality, and ownership.
Get Legal Guidance on AI Content in Dallas, Texas
The intersection of AI and intellectual property law is evolving rapidly, bringing new opportunities—but also legal uncertainty. As courts begin to address foundational questions about AI training, output, and fair use, it’s essential for content creators, developers, and rights holders to stay informed. Monitoring updates from the U.S. Copyright Office, proposed legislation, and credible legal analysis will be key to navigating this shifting terrain. To protect your rights and assess your exposure, consulting with an experienced IP attorney can help you build a tailored strategy aligned with current law and emerging trends. At Griffith Barbee, our attorneys regularly publish insights on these topics and provide practical tools for creative professionals and businesses.
By understanding the applicable legal framework, maintaining clear records of human involvement, and proactively evaluating IP risks, creators and companies can harness AI technology with confidence, while safeguarding their valuable intellectual property. If you have questions about ownership, licensing, or enforcement related to AI-generated content, contact our IP attorneys at Griffith Barbee PLLC in Dallas, Texas, for strategic legal counsel.