• Pascal's Chatbot Q&As
  • Posts
  • Disney and Universal’s lawsuit against Midjourney is not just a copyright dispute—it is a high-profile attempt to draw a legal and cultural line in the sand.

Disney and Universal’s lawsuit against Midjourney is not just a copyright dispute—it is a high-profile attempt to draw a legal and cultural line in the sand.

The complaint’s breadth, evidence, and tone all suggest a confident, well-prepared offensive that could reshape the obligations of AI image generators.


Disney and Universal v. Midjourney: Key Grievances, Legal Strengths, and Broader Implications for AI Companies

by ChatGPT-4o

Introduction

The lawsuit filed jointly by Disney and Universal against AI company Midjourney marks a pivotal moment in the evolving legal landscape around generative AI and copyright. Filed in the Central District of California on June 11, 2025, the complaint accuses Midjourney of direct and secondary copyright infringement for generating and publicly displaying unauthorized images of iconic characters from the plaintiffs’ vast IP catalogs—including Darth Vader, Elsa, Shrek, Homer Simpson, and the Minions. This is the first time that major Hollywood studios have taken direct legal action against an AI image-generation firm, and the implications could be wide-reaching for the AI industry.

Key Grievances in the Complaint

The plaintiffs assert several core grievances, which are structured around Midjourney’s alleged willful disregard for copyright protections:

  1. Systematic and Willful Infringement
    Midjourney is accused of “reaping the rewards” of decades of creative and financial investment by Disney and Universal without permission or compensation. The plaintiffs describe Midjourney as a “virtual vending machine” for high-quality infringing images generated at scale.

  2. Unauthorized Use of Copyrighted Characters
    The complaint includes explicit examples of Midjourney users generating images of characters like Darth Vader and the Minions simply by typing prompts. Midjourney's outputs are alleged to be high-fidelity visual replicas derived from training data that includes plaintiffs' copyrighted works.

  3. Promotion and Commercial Exploitation
    Midjourney is accused of promoting infringing outputs on its platform—using the plaintiffs’ characters to market its services, implying a false endorsement. The complaint also describes a tiered subscription system that monetizes this infringing activity, earning Midjourney $300 million in 2024 alone.

  4. Failure to Implement Reasonable Safeguards
    Plaintiffs allege that Midjourney has refused to adopt standard copyright protection measures, such as prompt filtering or output screening, even though similar controls are widely used to restrict nudity or violence. According to the complaint, the plaintiffs asked Midjourney to implement such measures, but their requests were ignored.

  5. Imminent Expansion to Video Infringement
    The lawsuit emphasizes that Midjourney is preparing a generative video tool trained on the same copyrighted material. Plaintiffs argue that this exacerbates the risk of infringement and demands urgent injunctive relief.

  6. Injury to Copyright Incentives and the Creative Economy
    The studios argue that Midjourney’s actions undermine the incentives foundational to U.S. copyright law—citing economic data that the U.S. entertainment industry supports 2.3 million jobs and contributes $260 billion to the economy.

Quality and Strength of the Evidence

The complaint is methodically structured, leveraging detailed factual assertions alongside legal arguments. Its strengths include:

  • Extensive Character Descriptions
    Plaintiffs describe the physical and conceptual qualities of copyrighted characters—Buzz Lightyear’s spacesuit, Hulk’s green musculature, Shrek’s trumpet ears—making a compelling case that these elements are copyright-protected expressions rather than general ideas.

  • Concrete Examples of Infringing Outputs
    The complaint contains visual comparisons and descriptions of infringing images generated by Midjourney, supported by publicly visible outputs and metadata.

  • Business Model Context
    Plaintiffs outline how Midjourney monetizes this infringement—via subscriptions, cloud GPU processing, and promotion—undermining any potential fair use defense based on noncommercial or transformative purposes.

  • Pre-Litigation Conduct
    The inclusion of cease-and-desist letters sent by Disney and Universal, which Midjourney either ignored or dismissed, strengthens the claim of willful infringement.

However, some areas could be vulnerable:

  • Training Data Evidence Is Not Fully Detailed
    The complaint does not provide a technical breakdown of how copyrighted works were used in training datasets—something that will likely be contested in discovery.

  • No Evidence Yet of Video Outputs
    Although the complaint raises concerns about Midjourney’s forthcoming video service, it lacks current examples of infringing videos, which could weaken preliminary injunctive relief related to video.

Most Surprising, Controversial, and Valuable Statements

Surprising:

  • Midjourney’s revenue reportedly surged from $50 million in 2022 to $300 million in 2024, demonstrating a massive commercial scale to the alleged infringement.

  • Plaintiffs include characters from virtually every major Disney and Universal franchise—Marvel, Star Wars, Pixar, DreamWorks, and The Simpsons—suggesting a sweeping infringement claim of unprecedented breadth.

Controversial:

  • Plaintiffs call Midjourney a “bottomless pit of plagiarism,” invoking moral condemnation beyond legal argumentation.

  • The suit critiques Midjourney's CEO for comments implying that AI tools could be used to “cut artists out,” framing the issue as a labor and cultural rights threat.

Valuable:

  • The case could set precedent on whether training generative AI models using copyrighted visual materials constitutes infringement.

  • It emphasizes that AI companies can no longer shield themselves by pointing to technical distance from user prompts or blaming training-data aggregators.

Assessment of the Plaintiffs’ Chances of Success

Disney and Universal present a highly compelling complaint, bolstered by:

  • Direct Output Evidence: Showing near-identical replications of copyrighted characters;

  • Clear Economic Harm: Given the direct monetization of infringing content and competition with licensed merchandise;

  • Willfulness and Notice: Midjourney ignored cease-and-desist letters, increasing liability risk.

However, potential defenses could include:

  • Fair Use: Midjourney may argue its outputs are transformative or generated by users, not the company.

  • Lack of Direct Copying: Midjourney may claim it did not store or distribute exact copies, instead generating new outputs based on latent space representations.

  • Section 230-Like Shielding: Though not a strong defense here, Midjourney might analogize itself to a neutral platform.

The court may initially grant partial injunctive relief, particularly against promotional use or commercial expansion into video. Damages could be substantial if discovery reveals large-scale use of registered copyrights.

Implications for Other AI Cases and Industry Practices

This lawsuit may catalyze:

  1. Tighter Scrutiny on Model Training Data
    Plaintiffs will likely push courts to define whether ingesting copyrighted data into training models constitutes “copying.”

  2. Pressure on Smaller AI Startups
    If the court favors the plaintiffs, other startups relying on non-licensed datasets could face similar suits or preemptive settlements.

  3. Need for Copyright Safeguards and Licensing
    AI developers must begin building copyright protection into the architecture of their tools—through prompt filtering, dataset cleansing, and provenance tracking.

  4. Shift Toward Licensing Deals
    Much like the New York Times’ suit against OpenAI, this case may accelerate licensing negotiations between studios, publishers, and AI firms.

Recommendations for AI Makers

  • Implement Prompt and Output Filters: Proactively block prompts involving protected IP, following approaches used by OpenAI and Adobe Firefly.

  • Disclose Training Sources Transparently: Voluntarily reveal datasets and secure licenses or remove unlicensed materials.

  • Develop Opt-Out/Opt-In Mechanisms: Allow rights holders to exclude their content, particularly ahead of legal mandates.

  • Negotiate Licensing Partnerships: Engage studios, labels, and publishers in good-faith content access deals.

Conclusion

Disney and Universal’s lawsuit against Midjourney is not just a copyright dispute—it is a high-profile attempt to draw a legal and cultural line in the sand. The complaint’s breadth, evidence, and tone all suggest a confident, well-prepared offensive that could reshape the obligations of AI image generators. While Midjourney may mount technical or fair use defenses, the balance appears to tip toward the plaintiffs. If successful, the case will send an unambiguous message: ignoring copyright in the age of generative AI is not just risky—it may be ruinous.