.st0{fill:#FFFFFF;}

Disney and Universal Sue Midjourney: Is Generative AI Just High-Tech Piracy in Disguise? 

 June 17, 2025

By  Joe Habscheid

Summary: Disney and Universal have launched a high-stakes legal battle against the AI startup Midjourney. The claim? Systematic copyright violations that cut to the core of how generative AI is reshaping—and possibly undermining—creative industries. This lawsuit doesn’t just target a few bad outputs. It challenges the model's entire pipeline and assumes a frontal assault on its training data, commercial intent, and defense strategies built on "fair use."


Midjourney’s Accusers: Copyright Powerhouses

When companies like Disney and Universal step into the courtroom, it’s not a tremor—it’s an earthquake. These aren’t startups testing the legal waters; they’re corporate giants who’ve spent decades building global media empires on intellectual property. Now, they’re pointing fingers directly at Midjourney, asserting the AI firm has engaged in massive-scale infringement of visual assets.

They call Midjourney a "bottomless pit of plagiarism," a phrase chosen not for drama but to frame Midjourney’s business model as inherently illegal. The studios argue that Midjourney doesn't merely copy—it mass-produces unauthorized reinterpretations that dilute the market and threaten existing rights holders.

The Key Evidence: Familiar Faces in AI-Spawned Outputs

The complaint submitted by Disney and Universal includes dozens of visual outputs that allegedly demonstrate Midjourney’s ability to recreate iconic characters and scenes. That’s not a casual slip-up in a sea of content. It’s a curated takedown: images resembling Yoda wielding a lightsaber and characters that look suspiciously like The Boss Baby.

This move strategically shifts the spotlight from the AI’s mechanics to its results. It asks a simple question: If the output walks and talks like copyrighted content, doesn’t that make Midjourney liable?

Legal Positioning: Fair Use Under Fire

Chad Hummel, a known intellectual property attorney, frames the body of visual evidence against Midjourney as particularly damning. His assertion? These outputs are “not sufficiently transformative,” a vital legal benchmark when determining whether a secondary use qualifies as fair use.

In simpler terms: If your generated image looks like what someone else already made—and doesn’t dramatically alter its meaning, expression, or value—you might not get the court’s blessing under fair use. And according to Hummel, that’s the trap Midjourney has likely stepped into: familiar characters, familiar settings, delivered through a different tool but with the same purpose.

Shifting Legal Ground: Why This Case Feels Different

For Emory law professor Matthew Sag, this lawsuit isn’t just another skirmish between creativity and code. He notes that prior AI copyright disputes leaned on “cherry-picked” outputs to argue harm. This suit is broader and arguably more focused. It doesn’t aim cherries—it goes after the orchard.

The plaintiffs accuse Midjourney of training its models on copyrighted materials—scraped en masse from the open internet—and creating not only derivative artworks but also duplicating content during the model's very formation.

Did Midjourney Ignore Warnings?

According to the lawsuit, Disney and Universal didn’t arrive at court as their first line of defense. They claim they requested Midjourney to adopt tech safeguards that would prevent the output of infringing content. The answer, according to the studios? Silence—or worse: dismissal.

That kind of narrative builds a stronger plaintiff story. Courts don’t like defendants who ignore requests for moderation. If Midjourney did nothing after being told about violations, it strengthens the studios' framing that the model’s behavior is not just harmful—it’s deliberate.

The Training Data Question: Cleaned or Complicit?

One of the most controversial claims in the lawsuit is about how Midjourney prepared its training data. The complaint alleges that Disney and Universal works weren’t just passively absorbed—they were “cleaned.” That means someone at Midjourney allegedly took steps to isolate and structure IP-laden material for training.

Cleaning copyrighted data doesn’t remove the violation. It may, in fact, underscore it—because it suggests someone knew the risks but still took intentional steps to fold the material into the training mix. That makes this less about ignorance, more about willingness to gamble on legal ambiguity.

What Disney and Universal Want the Court to Believe

This case isn’t just about Midjourney making pretty pictures. The lawsuit frames the AI firm’s strategy as an assault on creative labor and commercial IP. Disney’s general counsel, Horacio Gutierrez, is blunt: “Piracy is piracy,” whether committed by a teenager in a sock drawer or a billion-dollar AI startup in San Francisco.

Their message to the court hinges on a sense of urgency and alarm. These aren’t just business losses—they’re existential risks. According to the studios, Midjourney’s model isn’t an innocent byproduct of innovation—it’s industrial-grade bootlegging, dressed in Python code and GPU cycles.

Framing the Stakes: $260 Billion and Millions of Jobs

Here’s where the lawsuit widens its scope. The studios claim this legal fight isn’t about just two companies—it’s a fight to protect the entire U.S. film industry. They peg its economic value at over $260 billion and cite millions of jobs that are potentially at risk.

This is a classic social proof strategy. If the court doesn’t care about studio profits, perhaps it will care about a broader economic impact. And in a country where jobs often carry more political leverage than copyrights, it’s a tactic designed for more than the courtroom—it’s made to win over public opinion.

Implications: A Tipping Point for Generative AI?

If the studios win, it would rewrite the rules for how AI developers source, use, and process training data. This case will likely test how courts interpret the threshold between transformation and theft. More importantly, it will shape whether generative AI can function as it currently does or must become far more restricted.

And here's the critical question developers everywhere should ask: What happens to innovation when the cost of licensing outweighs the reward of creativity?


Conclusion: Midjourney’s legal clash with Disney and Universal goes far beyond celebrity characters and AI hype. It taps into a root issue at the heart of every disruptive technology: Who truly owns the inputs and outputs of machine learning? The studios have drawn their line in the sand. Now the AI world must decide whether to cross it—or build a new model entirely.

#AIlegalBattle #CopyrightInfringement #DisneyVsMidjourney #GenerativeAIEthics #IPRights #FairUseDebate #AIandCopyright #CreativeIndustriesUnderThreat #TechPolicy

More Info -- Click Here

Featured Image courtesy of Unsplash and Claudio Schwarz (0cAuOkTYVpo)

Joe Habscheid


Joe Habscheid is the founder of midmichiganai.com. A trilingual speaker fluent in Luxemburgese, German, and English, he grew up in Germany near Luxembourg. After obtaining a Master's in Physics in Germany, he moved to the U.S. and built a successful electronics manufacturing office. With an MBA and over 20 years of expertise transforming several small businesses into multi-seven-figure successes, Joe believes in using time wisely. His approach to consulting helps clients increase revenue and execute growth strategies. Joe's writings offer valuable insights into AI, marketing, politics, and general interests.

Interested in Learning More Stuff?

Join The Online Community Of Others And Contribute!