Disney sues Midjourney


log in or register to remove this ad


This was a couple of weeks back--I meant to post about it then.


But yeah, as I've been saying for ages, the people who stop the AI scraping of copyrighted works isn't us, it's the Disneys of the world, via the court system. When the AIs are scraping Star Wars and Mickey Mouse and stuff, that's when they'll step in. And they're the ones with the clout and the power to actually punch these AI companies. Guess they scraped the wrong thing!

The suit describes the AI companies as "a bottomless pit of plagiarism". They say "Piracy is piracy, and the fact that it's done by an AI company does not make it any less infringing".

It's not just Disney, either. Universal, too.

This will happen more and more. Once the big IP holders start to feel threatened, they're gonna start stomping on the AI companies. Like this.
 



This just reminds me that Disney (not the artist themselves) own the copyrights and could if they wanted to train an AI on what they already have.
I don't know who the artist would be for infringing depictions of Darth Vader or Iron Man, but it's not a single person. The AIs scrape millions of images.
 

As the video points out, at this stage the many lawsuits around the world currently focus on the output--the infringing images--rather than the piracy involved in obtaining those images.

When it came out recently that Meta had torrented -- yeah, torrented, like how a kid in his bedroom torrents games and movies -- millions of novels to feed into its Ai, the issue of piracy came into stark relief. They didn't even have the good manners to buy the books they were scraping--they stole them.

So while these lawsuits are about the infringing output -- "you're profiting off images of Darth Vader" -- I imagine that we'll soon start to see suits revolving around the actual act of piracy involved in obtaining that data in the first place.

Whatever happens, the AI companies are going to find themselves in constant legal battles going forward.
 


It is an interesting question. Generative AI is an unambiguously powerful and useful technology. But it requires a huge amount of data to train. So, how do we (as a society, I mean) build it ethically?

You see lots of people just dismiss the technology completely based on how it was created, but that doesn't answer the bigger, frankly more important questions.
 


Pets & Sidekicks

Remove ads

Top