AI-generated images of Superman, Batman and Bugs Bunny now sit at the centre of a courtroom drama that may reshape how Asian markets think about intellectual property in the age of generative AI.
Warner Bros. Discovery has sued Midjourney, alleging systematic copyright infringement through AI-generated images of iconic characters.,The case follows similar lawsuits by Disney and Universal, signalling a broader entertainment industry pushback.,The outcome could set new precedents across Asia, where creators, tech firms and regulators are wrestling with AI’s legal and ethical boundaries.
A lawsuit with capes and cartoon rabbits
The focus keyphrase here is Warner Bros sues Midjourney, and the story reads more like a plot twist than a business dispute. Filed in Los Angeles federal court, Warner Bros. Discovery (WBD) claims Midjourney has “brazenly dispensed” its intellectual property, training AI models on unauthorised copies of films, comics and animations. Among the characters reportedly cloned without consent are Superman, Batman, Wonder Woman, Bugs Bunny and Scooby-Doo. In short, the entire pantheon of DC and Hanna-Barbera has found itself reincarnated in AI art feeds.
WBD alleges Midjourney has created a commercial service that enables users to remix, reimagine and download derivative works, from comic-book icons to Saturday morning cartoons. What makes this particularly combustible is that Midjourney had briefly restricted such content when it launched video generation, only to later reverse course and market the change as an “improvement.”
The stakes in dollars and precedent
The 87-page complaint does not just request damages. It seeks a permanent injunction, statutory damages of up to US$150,000 per infringed work, and disgorgement of Midjourney’s profits. The claim rests on the argument that Midjourney’s infringement is not accidental but “systematic, ongoing and wilful.”
For Warner Bros, this is about more than money. It is about protecting cultural icons that underpin billions in box office, licensing, and merchandise. For Midjourney, the risk is existential. If the court rules against it, the service could face crippling payouts, reputational damage, and pressure to radically re-engineer its product.
The Disney and Universal chorus
This lawsuit arrives only months after Disney and Comcast’s Universal filed their own complaint. Their list of allegedly infringed icons reads like a pop culture encyclopaedia: Darth Vader, Baby Yoda, Homer Simpson, Spider-Man, Ariel from The Little Mermaid, and characters from Toy Story and Shrek.
Disney and NBCU were blunt: “Piracy is piracy, and whether an infringing image or video is made with AI or another technology does not make it any less infringing.”
The resonance is clear. Hollywood’s biggest names are drawing a line in the sand. Generative AI platforms, they argue, cannot simply absorb decades of creative output, spit it back out as novelty content, and sell subscriptions on the back of it.
Implications for Asia’s creative economy
Why does this matter for Asia? The region has both voracious consumers of Western franchises and its own burgeoning IP empires, from anime in Japan to K-pop-driven visuals in South Korea and an expanding animation industry in China. If Warner Bros succeeds, Asian rights holders will be emboldened to challenge local and global AI players.
Equally, regulators from Singapore to Seoul are closely watching how the US handles these disputes. As countries in Asia draft AI frameworks, rulings like this one may guide how to balance innovation with copyright protection.
For example, Singapore’s Infocomm Media Development Authority (IMDA) has already signalled that “responsible AI” includes respecting intellectual property. Meanwhile, Japan has been debating how copyright applies to machine learning datasets, especially given its huge manga industry. A comprehensive overview of how countries are approaching AI regulation can be found in this report on AI governance by the OECD.
A moment of reckoning for AI platforms
The case poses uncomfortable questions. Should AI models be trained on copyrighted works without explicit consent? If yes, where does fair use end and infringement begin? And if no, what happens to the pace of AI development, given that so much digital culture is copyrighted?
For Midjourney, its early decision to block outputs featuring Warner Bros characters suggests awareness of the risks. Its subsequent reversal, framed as a user-friendly upgrade, may now serve as a smoking gun in court.
The lawsuits piling up against Midjourney and other AI firms are more than corporate battles. They are shaping the rules of creative ownership in a world where anyone with a subscription can summon Superman into a brand-new universe. The result will reverberate through Asian tech ecosystems, where developers and studios are deciding whether to see AI as a collaborator, competitor, or a threat to decades of creative investment. The debate around AI's impact on creative industries is also explored in the context of AI Artists are Topping the Charts Weekly. This also highlights why When AI Slop Needs a Human Polish is becoming increasingly important.
The question is no longer whether AI can create compelling art. It is whether the legal and commercial structures around it can keep pace. The legal frameworks in Asia, such as Taiwan’s AI Law Is Quietly Redefining What “Responsible Innovation” Means, are also evolving to address these complex issues.






Latest Comments (4)
This situation with Warner Bros. and Midjourney echoes similar discussions here in Europe, en effet. We see the same challenges with copyright and training data, particularly concerning the EU's proposed AI Act and its provisions on transparency for foundation models. The idea that Midjourney "briefly restricted such content" then reversed course is quite telling, n'est-ce pas? It suggests a known vulnerability.
I'm wondering how Warner Bros plans to prove intent from Midjourney. Like, saying it's "systematic, ongoing and wilful" feels like a really high bar to clear, especially when Midjourney briefly restricted content and then changed its mind. It seems like that could be spun either way.
While the US courts battle over Superman and Bugs Bunny, it will be interesting to see how this informs the ASEAN Framework on Digital Data Governance. The argument that Midjourney’s infringement is “systematic, ongoing and wilful” could be a key point for future regional IP discussions, especially concerning AI model training data.
Counterpoint: I've seen some amazing AI art where people tweak these characters slightly. Does that count as "systematic" infringement if it's just a few pixels here and there?
Leave a Comment