Submitted 2 days ago by Beep@lemmus.org to technology@lemmy.world
https://jasonwillems.com/technology/2026/02/02/AI-Copyright/
Copyright became a problem for wealthy interests where it wasn’t before. So now they want exemptions for themselves only.
No You train your LLM on the public domain that’s available for everyone to use. If you don’t want your chatbot sounding like the 1920s, you should have thought of that before robbing humanity of a robust public domain.
AI very much did break Copyright law by taking stuff without having a license for it.
I haven’t read the article, but if the headline already starts out this wrong I don’t think it’ll get better.
The idea of copyright is to protect the financial rights of creatives, thus incentivising people to make more stuff, right?
Well even before AI, it wasn’t doing its job very well on that front. The only ones with the power and money to be able to leverage copyright to protect their rights are those who are already so powerful that they don’t need those protections — big music labels and the like. Individual creatives were already being fucked over by the system long before AI.
If you haven’t read the article, I’d encourage you to give it a try. Or perhaps this one, which goes into depth on the intrinsic tensions within copyright law.
You probably should’ve because yeah, the way AI companies are testing creative works is disgusting and downright wrong, but copyright law has very much been broken ever since the Internet because a thing. It’s just silly to treat works published on the internet the same way you treat books, paintings and DVDs.
Why tie it to death, why not plain 5 years?
If it was the day after they died, mightn’t that have an unintended consequence of making it more likely that copyright holders would start “falling out of windows” just when it’s convenient for producers and AI crooks to snaffle up their content, royalty-free?
So torrents don’t violate copyright law, now?
The people that create and run them broke copyright law.
If they do it's not by the actual training of AI.
It doesn’t break copyright laws because training something on any kind of data, as long as the data was legally obtained is legal (this includes scrapping publicly available data).
You can’t generate a sonic picture and sell it for the same reason you can’t draw sonic in Photoshop and sell it. These are tools and it’s up to the user to use them in a legal way.
Fan art is actually illegal, companies let it be because they get instantly thrashed by fans if they complain.
It's weird how AI has turned so much of the internet from its generally anti-copyright stance. I've seen threads in piracy and datahoarding communities that were riddled with "won't someone please think of the copyright!" Posts raging about how awful AI was.
I maintain the same view I always have. Copyright is indeed broken, because of how overly restrictive and expansive it has become. Most people long ago lost sight of what it's actually for.
The AI topic is botted massively, almost as much as political topics.
I don’t know who would benefit from a large portion of the Western youth being made to be disinterested in this emerging technology, but it isn’t the Western economies.
“Copyright” is an overloaded word that can both mean “IP/copyright law, it’s terms and enforcement” as well as “the rights of an author to decide how their work should be used”
In the vein that LLM are just a tool? Wouldn’t it be legally a problem if a photoshop filter had filters specifically to generate Sonic art?
Btw, why is that blue hedgehog still a thing?
Fan art is not illegal. As long as you’re not selling it, it would generally fall under fair use.
Fan art is not illegal.
Off the top of my head, I know both Disney and Nintendo have sued people for making fan art. Fair use doesn’t explicitly allow you to make fan art, regardless of its transformative nature, and whether or not you owe Disney hundreds of thousands of dollars for drawing Mickey depends on court review on a case-by-case basis because it not technically legal in the US. It may also not be technically illegal, but that doesn’t mean a corporation can’t sue you and be awarded millions in civil damages if they think you profited off the art in some way.
A quick google search will source you lawyers saying such.
…chicagobar.org/…/the-fine-line-between-fan-art-f…
It’s copyright infringements but like I said, most won’t bother fans not making a dime. There’s economic advantages to having fans create and distribute your content for free. A company can choose to copyright strike anything with their characters in it at anytime.
Ships already sailed but shouldn’t AI generated things just be considered derivative of their training set?
I don’t know how that works for images/video, but for code that means if it’s trained on GPL code the resulting code would have to be GPL and the liability is on the people distiributing AI generated code.
This also makes commerical use of AI generated anything complicated while allowing personal use under the current state of non-enforceablity.
Just like Aaron Swartz
I wouldn’t compare Swartz with the AI scrapers. Aaron pulled mostly public domain documents from JSTOR and caused minor issues with the servers which is “a bit” different than pulling everything from the internet to a database over a practically global DDOS-attack. But when companies do it it’s apparently somehow different and Swartz was pretty much publicly lynched and eventually bullied to suicide.
user28282912@piefed.social 2 days ago
The content produced by humans was scraped en-masse for the explicit purpose of training models which were then monetized into business products.
I struggle to reconcile that with Fair Use.
I can see if the source was EULA’d to remove all rights to what you post to things like Reddit, Stack Overflow, and if somehow those entities were contacted ahead of time and negotiated usage. You, I and the web server logs know that this was almost never the case.
Postimo@lemmy.zip 2 days ago
I think the core of the fair use argument is that the AI models that are being trained are transformative products of the original works.
Might be a hot take here but I basically agree. I still believe it was theft and that the realities of the legal framework we had don’t really stand up to the evolving problems, but under the current laws there is really no justification for saying that, taking the input of a bunch of images and giving the output of a set of statistical correlations of pixels based on descriptions, isn’t transformation.
red_bull_of_juarez@lemmy.dbzer0.com 1 day ago
I agree. I think that all AI companies need to crash and burn. But it’d be disingenuous to claim that what these models are doing is completely different than what humans are doing. Humans don’t pull stuff out of thin air. We are products of our upbringing and schooling. I say that, because I hate our current copyright laws with a burning passion and have done so long before LLMs showed up. It’s possible to hate copyright and AI companies.