Disney And Comcast's AI Lawsuit May Open A Pandora’s Box

The ongoing debate over generative artificial intelligence — particularly when such use conflicts with trademarks and copyright law — came to a head last week in what could be one of the most significant lawsuits in determining the future of the entertainment industry. The co-plaintiffs are two of the most powerful companies in the business: the Walt Disney Co. and Comcast Corp. The defendant is Midjourney Inc., a generative AI startup that the media giants are accusing of copyright infringement.

The lawsuit could prove a fascinating feat of Hollywood needle-threading, in which these powerful companies attempt to separate unacceptable and acceptable generative AI. It’s a distinction that appears mostly rooted in what costs them money versus what saves it.

To understand what’s at play, it’s important to clarify what Midjourney does. Its software (available via the company’s Discord and website) allows users to input prompts that create AI-generated images. Unsurprisingly, the results often include broadly recognizable pop-culture creations that are protected by copyright. This includes Disney’s Mickey Mouse, Marvel, Simpsons and Star Wars characters, and Comcast favorites such as the Minions and Shrek. The lawsuit alleges that as Midjourney has built its models from datasets that have scraped images from across the internet without permission and has ignored requests to cease doing so, it should have to pay $150,000 per infringement.

The odds don’t look good for the startup. Side-by-side comparisons of supposedly user-suggested, AI-generated frames are all but indistinguishable from simple screenshots from the properties in question, similarities that suggest copyright-infringing use of the source materials.

It sounds like a pretty simple, open-and-shut case. But what makes it all complicated and noteworthy is that this suit arrives at a time of potential upheaval. This is a tricky and tenuous moment in which much of the entertainment industry — including Disney and Comcast — is testing the waters with generative AI amid questions of ethics from the perspectives of both labor and copyright.

Disney itself, for example, retained the services of Method Studios (after soliciting several AI vendors) to create the opening credit sequence of the 2023 Disney+ Marvel series Secret Invasion using AI. Disney took pains to clarify exactly how it used the service in a statement to the Hollywood Reporter at the time: “AI is just one tool among the array of tool sets our artists used. No artists’ jobs were replaced by incorporating these new tools; instead, they complemented and assisted our creative teams.”

This kind of clarification and explanation is necessary because the entire subject of generative AI — a topic that is vast and often misunderstood or mischaracterized — is so loaded in the industry at the moment. Creative people of all stripes, from visual artists to craftspeople to screenwriters to actors, are envisioning worst-case scenarios for computer models putting them out of work. And it’s worth noting that the Midjourney suit does not address that controversial aspect of such software, or at least not directly. That’s probably because film studios are already using AI for preproduction development and special effects, which are pieces of the filmmaking puzzle that previously employed visual effects artists and companies.

Disney’s Chief Legal Officer Horacio Gutierrez has confirmed, in his statement about the lawsuit, that the company is “bullish on the promise of AI technology and optimistic about how it can be used responsibly as a tool to further human creativity.” So, this particular legal battle seemingly has a much more direct aim. “This is not going to be Hollywood trying to shut down generative AI,” Chad Hummel, principal at the Los Angeles office of the law firm McKool Smith, told the Washington Post. “It’s about compensation.”

It’s hard to imagine Midjourney mounting an effective defense against those claims. As multiple reports have pointed out, the company’s CEO David Holz admitted in a 2022 Forbes interview that Midjourney built its dataset by performing “just a big scrape of the internet,” and made no attempt to seek consent from artists or copyright holders. “There isn’t really a way to get a hundred million images and know where they’re coming from,” Holz said.

It’s a damning admission, yes — but one that’s echoed across the generative AI landscape, where software creators such as OpenAI insist that “it would be impossible to train today’s leading AI models without using copyrighted materials.” This could make Disney and Comcast’s lawsuit, where they call Midjourney’s AI-powered image generator a “bottomless pit of plagiarism,” even more striking. The plaintiffs could be opening a Pandora’s Box since “the promise of AI technology” in other realms, such as screenwriting and special effects, would presumably require dipping into that same “bottomless pit.”

Hollywood studios will aim to draw a line in the (quick)sand by throwing around buzz phrases such as “responsible use” or even that twistiest of vernacular pretzels, “ethical AI”— a phrase which, in its oxymoronic implications, is something akin to “jumbo shrimp.”

We’re often told that it’s too late to put the genie back in the bottle, and that might be true. But when studios attempt to rein in the same technology that they’re eagerly using elsewhere, the more urgent question isn’t about retreat. It’s whether Dr. Frankenstein can control the monster he’s already unleashed on the world.