Disney vs Midjourney.

I have recently been made aware of a lawsuit filed by Disney against Midjourney, predictably over alleged copyright infringement. The core of the accusation lies in the fact that one can use the Midjourney AI to generate images depicting well-known characters that are considered the intellectual property of Disney.

Naturally, we are speaking of iconic figures such as Mickey Mouse, Elsa, or even Kung Fu Panda. (Though the latter originates from DreamWorks, it is often lumped into broader pop-cultural references, further muddying the waters of authorship and ownership.) However, there is a certain irony to be found in the notion of Disney claiming absolute ownership over characters whose roots, in some cases, are far older than the company itself.

Take Snow White, for example. Her tale can be traced back to the Brothers Grimm, whose 1812 collection Kinder- und Hausmärchen formally introduced the character into European literary canon. Similarly, Hansel and Gretel also appear in the same collection, their origins stemming from centuries of folk narratives passed down through oral tradition. These stories existed long before Disney ever adapted them, transforming public domain material into profit-generating intellectual property. (The practice of commercialising folklore is hardly new, but it was Disney who perfected it into a global strategy throughout the 20th century.)

At the heart of the lawsuit lies the allegation that Midjourney is actively enabling — or at the very least, passively allowing without meaningful intervention — the creation of images that replicate characters owned by Disney. These images, according to the company, are said to cause financial harm.

Yet before we can assess whether such harm is genuine or merely perceived, we must first take a closer look at Disney’s actual business model. One must ask: where does the House of Mouse truly make its money?

Let us take the example of Frozen, and more specifically the character of Elsa. The film itself grossed approximately 1.3 billion USD at the global box office upon its release in 2013 — a figure impressive by any cinematic standard. However, what is often overlooked is that this represents only a fraction of the true economic engine behind the franchise. According to a 2014 report by Forbes, merchandise sales tied to Frozen — from lunchboxes and pencil cases to backpacks, costumes, and even branded yoghurt — were estimated to exceed 5 billion USD in just the first two years following the film’s release.

When your child insists on an Elsa-themed pencil case, or refuses to accept a pencil sharpener unless it’s shaped like Mickey Mouse, or when a school backpack must bear the image of Darth Vader, rest assured: someone, somewhere, has paid a licensing fee to Disney. It is in these small yet ubiquitous consumer transactions that the true magnitude of the company’s revenue model is revealed.

I refer to this strategy as the Hello Kitty model — the transformation of fictional characters into lifestyle accessories, omnipresent in the daily rituals of consumption. In this framework, the films themselves serve merely as elaborate commercials — narrative vehicles designed to ignite emotional attachment and brand loyalty, which are then monetised through merchandise. From this perspective, the financial relevance of the film Frozen becomes marginal when compared to the empire of products it has spawned. The proportion of revenue derived from merchandising dwarfs that of the box office, rendering the cinematic work itself almost incidental in economic terms.

Let us be perfectly clear: with this lawsuit, Disney is not attempting to protect the work itself — that is, the film Frozen — but rather the merchandising ecosystem built around it, which is vastly more lucrative.

By suing Midjourney for damages, Disney does not appear genuinely concerned that someone might use AI to recreate the narrative arc or animation of Frozen in its entirety. What it truly fears is the potential for AI to generate merchandise — particularly digital merchandise — that features characters whose aesthetic or branding falls within Disney’s domain.

And this concern is not entirely unfounded.

To grasp the scope of this, one might look outside the Disney universe for a moment. Consider the case of Warhammer 40,000, the iconic tabletop miniature wargame by Games Workshop. A single fully playable set of miniatures can easily cost upwards of $250, and that’s often just the entry point — seasoned players typically own multiple armies, terrains, and accessories. But then came the 3D printer revolution.

The introduction of affordable 3D printers and consumer-grade 3D scanners in the early 2010s disrupted this niche but high-margin market almost overnight. Enthusiasts were suddenly able to create remarkably convincing replicas at a fraction of the cost. Online repositories of 3D-printable models began circulating, and with them, the profit margins of physical miniature producers were — to put it politely — significantly compromised.

A case in point: miniature replicas of characters like Lieutenant Titus, shown below, are now widely available in formats compatible with home 3D printers. And while they may lack the official branding or minute detailing of licensed models, they are good enough for casual and even some competitive players.

titus

This is exactly the kind of disruption Disney sees looming on the horizon. Not the wholesale recreation of films, but the fragmentation of its control over the aesthetic capital it has so painstakingly cultivated. If an AI can produce “Frozen-esque” artwork or characters in the Disney style — and users can then convert that into printable or virtual merchandise — the traditional licensing model begins to crack.

Now that I have laid out the rational, economically grounded reason behind this dispute, you might be inclined to conclude that the lawsuit — or more broadly, this cultural and technological skirmish — makes sense. But that is not at all what I am saying.

The point of referencing the Titus miniatures was not to defend Disney’s stance, but rather to illustrate that the Midjourney episode represents merely a transitional phase — a precursor, not a conclusion. It is not the final form of the disruption we are witnessing. In fact, it’s already outdated compared to what is happening in the world of physical replicas.

Let me explain.

To generate your very own image of Elsa — perhaps in ways Disney would find profoundly objectionable (see: NSFW fan-made Elsa imagery) — one currently needs to access a platform like Midjourney or an equivalent AI image generator. But when it comes to something like the Titus figurine, the infrastructure for replication has already decentralised entirely.

With a halfway-decent 3D printer and a bit of time, people are manufacturing high-quality miniatures — and in many cases, selling them — right from their garages. There is no need for a corporate middleman, no need for licensing, and no real mechanism for enforcement. The genie is already out of the bottle.

Now let’s return to digital image generation.

At present, if you wish to run Stable Diffusion locally and generate highly specific characters — say, Elsa in an alternative setting — you would need to invest roughly €1,500 in a high-end GPU. You would also need to install the appropriate models, perhaps train or fine-tune them with examples, and acquire some minimal technical fluency. This sets a barrier of entry, but it is not a permanent one.

Because here is the crucial insight: modern professional laptops increasingly ship with competent GPUs, and some even include dedicated AI accelerator chips (like Apple's Neural Engine or Intel’s AI Boost). The hardware gap is shrinking — fast. What was once confined to industrial servers is now trickling down to consumer-grade hardware.

So we must ask ourselves: how much longer will the average user truly need platforms like Midjourney at all?

In my estimation: three, perhaps four more years — at most.

The moment high-performance generative models can be run directly on a standard laptop, and image training becomes semi-automated or commodified, the game changes. At that point, any regulatory strategy based on platform control becomes obsolete. Just as no one today can realistically stop people from 3D-printing a miniature of Lieutenant Titus, no one will be able to stop users from generating endless variants of Disney characters — or anything else — in the privacy of their own homes.

So, where does this lawsuit actually come from?

The root cause is — unsurprisingly — the shareholders.

We are all well aware by now that generative AI poses a fundamental challenge to traditional copyright law. We also know that when a company like Disney — which, let’s be honest, hasn’t exactly been enjoying a golden age financially (see: declining streaming margins, underwhelming box office returns, and activist investor pressure) — finds itself under threat from rapid technological change, the shareholders tend to ask themselves two very basic questions:

1) Should I consider investing in other, more future-proof companies? 2) Can we somehow stop technological progress?

Disney’s decision to sue Midjourney suggests that a significant portion of its executive leadership genuinely believes the answer to the second question might be “yes”. Or, at the very least, that it’s worth pretending so — for a little while.

This is not as far-fetched as it sounds. Corporate management teams are, as a rule, notoriously poor at understanding the trajectory of emerging technology. By the time a technology is on their radar, it's often already in widespread use by the general public. From their perspective, filing a lawsuit becomes less about actually halting the advance of AI, and more about sending a reassuring signal to investors: We’ve got this under control. Don’t sell your shares just yet.

In other words, the lawsuit is not really a line of defence against AI. It's more like a press release dressed up in legalese — a theatre of reassurance, staged for the benefit of the capital markets.

Will it succeed? That’s beside the point.

For companies like Midjourney, victory isn’t even necessary. All they really need is to delay the legal process — stretch it out just long enough, say three or four years — until we reach the inevitable technological tipping point: the moment when virtually every household computer comes equipped with a GPU powerful enough to run generative models locally.

At that stage, the landscape will have shifted entirely. By then, Midjourney might not even be producing still images at all. The focus may have moved on — to video generation, immersive 3D content, or something else entirely. And once that happens, this lawsuit — built entirely around the production and copyright of static images — will be rendered obsolete, a relic of a legal system chasing yesterday’s innovation.

In fact, the platform could evolve into a business that merely distributes finely tuned models, with users running them locally on their own machines — models, of course, available for purchase. And once that happens, neither Disney nor any court on Earth will be able to track, let alone regulate, what users are generating privately. The distribution becomes decentralised. The enforcement becomes impossible.

This is one of those lawsuits that only makes sense within the closed feedback loop of incompetent management trying to pacify skittish and uninformed investors. A sort of legal placebo, administered not to change the course of events, but to calm the nerves of people too slow to grasp what’s actually happening.

By the time the case reaches its conclusion — if it ever does — the world will have moved on.

This lawsuit is the very epitome of what happens when a company fails to acknowledge that a technological breakthrough has rendered its business model obsolete.

At such a crossroads, there are only two rational responses: either resist and deny the direction of progress, or embrace it and evolve. And the direction here is crystal clear — this kind of generative technology is on a path to become ubiquitous. It will live not only on powerful servers, but in every laptop, every smartphone, every device with a chip worth its silicon.

Disney, if it truly understood this trajectory, could in fact profit just as handsomely as Midjourney — perhaps even more so. Imagine, for a moment, a Disney-branded generative AI platform, one that operates within the boundaries of legality, where for a reasonable subscription fee, users could create their own fan art, posters, avatars, even short animations, using official Disney IP, all under licence. It would monetise engagement rather than criminalise it. And in doing so, it would meet the market exactly where it’s going.

But such a move requires one crucial ingredient: a management team that actually understands technology.

It requires vision, technical literacy, and above all, the humility to see disruption not as a threat, but as an opportunity.

And if this lawsuit is anything to go by, Disney’s current leadership clearly does not understand any of that.

Uriel Fanelli

Il blog e' visibile dal Fediverso facendo il follow a: @uriel@keinpfusch.net

Contatti: