US appeals court rules AI generated art cannot be copyrighted

It's not difficult to understand.

Replace the monkey with a 2nd human, and it's obvious that "the guy" does not earn the copyright, it goes to the person who took the photo. If there was no person, then there is no copyright.

The AI thing is no different. If I ask my human friend, "please paint a picture using your vast knowledge and experience", then my friend gets the copyright. Replace friend with AI; there is no person to assign the copyright, so there is no copyright. It doesn't default to me just because I asked for it.

Yeah I'm a little torn on this one. I generally think that much of IP law causes more harm than good, so in the abstract I'm in favor of copyright being weaker. But in this specific case, given the context of existing copyright law and its intent it seems pretty obvious to me that he should have copyright over the photo.

I don't think it's analogous to AI art though - no other humans creative input and therefore livelihood was ever involved in the process, and it's not like monkeys have any use for money or ownership of intellectual property. (Although the hypothetical situation where you assign the monkeys personhood and give them a bunch of royalties to pay for a better habitat and piles of bananas would be pretty cool.)

As I understand that is a misunderstanding of the case. They argued that the animal should get the copyright, and lost, because animals do not qualify. They did not establish that pressing the button is required for the human to qualify for copyright. They established that a monkey pressing the button doesn't qualify the monkey. (because the monkey never qualifies)

If they would have argued that the human should have got copyright for it, they almost certainly would have agreed. It's just, that wasn't the case they put forth.

It's worth pointing out that this was just a US Copyright Office ruling. It never went to court[1], where the "expert consensus" is that the photographer would have prevailed. But the value of the handful of photographs was tiny in comparison with the publicity (which was always true) so no one ever went to court to try to prove it.

It's not really clear to me how much this AI case matches though. There seems naively to have been a lot more creative work rigging that specific bit of monkey art than there is in applying a decidedly generic AI image generation tool. That AI is so much more capable as a machine for generating art than a camera is seems to cut strongly against the idea here.

[1] Note that PETA then tried to use this case to drive the converse point, suing on behalf of the monkey who they wanted to hold the copyright. They lost, unsurprisingly.

I see this as "That thing which doesn't work is currently not working. Again." The DMCA and copyright laws and regulations in the US are predatory nonsense, carefully crafted by lawyers in order to exploit the maximum amount of cash possible from people who actually do produce things.

The DMCA doesn't support artists and creators even indirectly; it empowers those least deserving and most ruthless to steal the profit, pat themselves on the back, and moralize about "following the law" to everyone else.

Copyright should be implicit and ironclad for 5 years. After that, 99.999% of sales have been made, whether your material is digital or otherwise. From 5 to 20 years, you should retain right to profits from the sale of any copy, but it should be 100% legal to copy, distribute, archive, remix, or whatever else you want with it so long as you aren't trying to sell it. After 20 years, public domain, no exceptions, no carveouts for family, friends, crafty lawyers, important politicians, or anyone else. No grandfathering, no special rules for special people.

Things made with AI should be protected by copyright, with the rights held by the user of the tool that generated the image. Like any other digital art.

There are machines that can paint your Dall-E renaissance creation onto a canvas with the style of your favorite master. The tools we have at hand have empowered us to rapidly and easily explore a vast domain of images, videos, music, voices, creative writing, and to do research and technical projects and write code in ways that were unthinkable 10 years ago.

These judges and lawyers think it's ok for them to rule on things without having the slightest clue as to the operation, function, and consequences of the technology - this ruling does nothing except to reinforce the status quo and empower the entrenched rights holders - the massive corporations, platforms, "studios", agents, and miscellaneous other gaggles of lawyers who trade in rights to media, but produce nothing of value in themselves.

Imagine a world in which content creators got paid a fair return relative to the revenue generated by their work, in which platforms and interlopers were limited to something like 5% of the total generated profit per work, after cost (to the creator). There'd be no incentive for bullshit rulings like this, with no angry mobs of litigious bastards with nothing better to do than sue for tampering with their racket. I cannot possibly see any other path to this ruling than this; else this judge is fortunate beyond words that his community has so uplifted the mentally deficient among them.

I agree - it's ridiculous. It's not much different to saying "you didn't take the picture; the actuator that opened and closed the aperture did".

Consider these (rhetorical, I am not sure I'm up for the nuanced debate given IANAL) questions:

1. Who owns the rights to a commissioned piece of art? The artist, or the commissioner? Which rights?

2. What about derived works of art made with or without the permission of the original artist(s)? When a book is turned into a film, who "rightfully" owns what? When the Rolling Stones wrote Sympathy For the Devil, did the estate of Mikhail Bulgakov have a right to feel aggrieved, and should they have received royalties?

3. What rights can be assigned/transferred, and what rights can't be? What needs to happen for that process to be legally binding?

4. Is a monkey capable of being a willing participant in a photograph, or a contract assigning rights in any way?

5. Same question, but for a machine? What does it mean for an AI to assign rights, or assert moral rights?

5. If the law makes it clear that a legal party to a statute (law), or contract must be a human or other legal subject (an incorporated business), can those laws and contracts lawfully apply to an animal or machine?

6. What is the intent of intellectual property law? Many argue it is mostly civil law, that follows the spirit of civil law in striving towards fairness?

We can argue if intellectual property law implementation is just, but your issue seems to be that the time invested in planning a creative act is the central tenet on which a copyright protection should be determined.

If so, Picasso was wrong to argue that his quick sketch on a napkin took him "a lifetime" to create, and your argument is just and correct. I disagree.

Regardless, what do you think the law is attempting to actually protect which is not "time taken to plan and create the work"?

Note when thinking about these questions it might be helpful to remember that ownership, copyright and moral rights are not all equivalent things in law.

It always felt to me like the photographer was trying to have it both ways there:

"Whoa! Isn't this sooo trippy! A monkey showing self-awareness to take a picture of itself!"

Courts: "Okay, the monkey took it, so no copyright for you."

"No, you don't get it! I put in a ton of work to stage that to the point that the monkey just had to be in the right place at the right time. Hell, a worm could have triggered it!"

I'd also think the creative input jumping mediums would also be a factor. Text to image is obviously a jump.

I think this specific quote from the article deals with this situation.

> U.S. Circuit Judge Patricia Millett wrote for a unanimous three-judge panel on Tuesday that U.S. copyright law "requires all work to be authored in the first instance by a human being."

You are right but there are a lot of curmudgeons that want you to get of their lawn with your AI. Really this whole situation is more of an indictment of copyright rather than of AI.

Maybe the best idea would be just to scrap copyright alltogether. It just blocks people from collaborating and building on top of each other's work. If everyone demanded royalties, where would Linux be? Wikipedia? scientific research? Could we even have this conversation in a forum?

It was insipred by these authors, not created. I won’t claim where copyright/authorship should be, but this reduction makes less sense than needed for important definitions.

> Wait just so i understand it, if a single human creates an AI model and trains it, and then prompts it to create an image, is that considered "human intervention" and does that make that human the author of that image?

No, you misunderstand. The human involved is explicitly claiming the work was entirely AI authored, and that it should be given a copyright registration with the AI as the author.

The human is not claiming that they should get a copyright as the author for the reasons you describe. Had the human claimed authorship, the results of the case might have been very different. This case seems to have been engineered to lose for publicity, rather than being a serious attempt to secure copyright on the work.

I think in later cases we'll get some tests to apply about how much human intervention is required.

Who trained the LLM is probably not the issue, the courts would likely want to know about the training material. If I trained a model exclusively on Warhol art, and then had that model create new images in Warhol's style, I didn't do any of the creative work and probably don't get the copyright. Warhol's estate probably owns the copyright to the model generated images as they are derivative works.

I do think that a model trained on many different artists' works, with me providing substantial feedback to the model (and I can show the process), probably will at some point give me the copyright.

Somewhere there is a line:

- "Make a picture of a mouse." Probably not giving you copyright

- Using a model to erase a powerline in a photograph you took. Probably you own the copyright to the original image and the one without a powerline in it (regardless of how many other people's images the model was trained on).

- "Make a picture of a mouse, who is bipedal, wearing pink shorts, with a chip in his ear, wearing sunglasses, with scruffy whiskers, holding a surfboard, on his way to the beach to hit some waves." then updating with "make him shorter, give him blue sneakers" and then updating with numerous other tweaks until you get it just the way you want. Who knows where this lands?

I think that in the short term the courts are going to land on the side of "anything made by a model trained on existing artwork is derivative of the training set so you can't own the copyright, no matter how much you tweak it." I think eventually the courts will recognize there is some amount of input that makes the computer image the realization of a vision in your head, and not a derivative of the training set. Just how every individual musical note has been played before, but at some point, you put them together in an arrangement that is original.

I also wonder this. I can write instructions to draw an image on the screen using OpenGL - or I can write an LLM and prompt it to draw an image. Why should I get authorship rights in one case but not the other?

> Wait just so i understand it, if a single human creates an AI model and trains it, and then prompts it to create an image, is that considered "human intervention" and does that make that human the author of that image?

I guess we will see when this gets tested in court. This current case linked to in the original article does not address this since the plaintiff already waived their own right to copyright already before copyright office.

There are 3 scenarios:

1) The AI should be the copyright holder (this judgement says NO).

2) If not 1 then the human should be the copyright holder via work-for-hire (this judgement says NO).

3) Human should be the copyright holder because they're the only human involved in the authoring (this lawsuit does not address this since direct copyright claims had already been waived).

I would assume that whomever prompts the AI is the author of the work. Adobe or Dell doesn't get to claim ownership to your work just because they made the tool or computer.

> if a single human creates an AI model and trains it

... on only their own artwork?

They are inspired though, not derivative. AI models contain no source data in a reproducible form (not that it really matters, but in case it is, they can’t).

>Thaler seems to go out of his way to claim no human intervention and authorship by the AI - So yeah, that's a very specific ruling that has little to do with AI as a tool. It's really more about AI personhood.

This, it was a poorly concocted scheme. People do stuff like this all the time, but even when they manage to confuse one branch of the government, the rest of the government isn't suddenly obligated to go along with it.

The prompt is a literary work independent of the system processing it. If the text is sufficiently elaborate, it is certainly copyright able. But the resulting image is still a different affair.

That sounds to me like a recipe. Recipes their interaction with copyright are well-established, legally speaking, in the United States.

You just stated how an AI generated image should be copyrightable. You should be able to own the copyright to all the configuration settings. If those settings then can be transformed to a 100% deterministic image (true, since you provide the seed) then I don't see how this is different than developing a photo negative film and transferring it to paper.

> This would mean I could copyright "man holding apple".

I think this is true today. You can have copyright on this phrase, just consider if it were the title of a song or poem.

You can click on "continue without supporting us".

+1

Copyrights are more about the process more than the product. This is why "clean-room" implementations do not violate copyright law.

Trademark is about commercial ambiguity.

Usually the method of a copyright troll never reaches the jury stage. It's mostly a racquet to get people to pay you not to sue them or file vaguely legitimate DMCA takedowns on their content.

It's amazing when laws make complete and consistent sense, +100 to this great answer

If I just send fake letters, it's illegal (I assume). If I have a legitimate website selling logos, and point to the product page for the logo I accused you of copying, and I can claim copyright ownership over AI generated art, then I have the law on my side even if I get taken to court (I assume).

I'm not a lawyer though, so I'm probably wrong. At the very least, the legitimate website makes the threatening letter look more believable.

I like this idea. It's like the library of babel (https://libraryofbabel.info), but for logos.

>Just set up a website that generates every combination of pixels as you scroll down it.

Sure. I guess when it finishes your great great grand-children (I might be very generous here too) can deal with the fallout of such a brute force algorithm.

I encourage you to read the opinion.

There was nothing to nail down here. The Copyright Office rejected the registration. The Review Board affirmed. The trial court affirmed. Three appeals court judges affirmed. No dissenting opinion.

Thank you. That sounds perfectly sensible.

> The court is ruling that computers don't have human rights.

No, it is just ruling that the Copyright Act requires human authorship. Whether computers have human rights is not an issue before the court.

you are right

But the LLM is a tool. If I use a set of colored pencils to draw you a bird, the pencil company doesn't own the copyright. I do. Because I used the tool.

So.. does the conductor of an orchestra get royalty rights? He's just prompting the "actual" musicians.

I've got a better, probably incoming situation:

I ask a deep-thinking LLM for a blog article, and to deliver that, it requests images from another LLM.

Everybody gangsta until an object object starts falling.

Unless you can make your prompt so specific that the AI generates substantially the same image every time you run it, I think you're perpetually vulnerable to the argument that significant decision making was done without human hands and therefore the work is not primarily human created.

> What I can guarantee, is that the prompt itself would be copyright-able.

That's non-obvious to me. Even if the prompt is extremely long and precise, if it is somehow purely functional, it seems possible for it to not be (although in practice, I agree that most prompts could be).

How about selection? If I select the good image from 1000 others? Curation is also a contribution to art.

If you want to know if this would be copyrightable, just flip a coin. I don't think anyone can give you better legal advice on this example than a coin toss.

https://en.wikipedia.org/wiki/Infinite_monkey_theorem

> The practical implication is you can't copyright something that your AI generated.

No, its not.

This is not a case of the human trying to claim copyright as the author of a work made using AI tools.

> As the article notes, copyright applications are also being rejected in cases where a human asserts authorship over an AI generated work.

That is true (although at least one has been accepted by the copyright office, IIRC), but it is not an outcome of this case (even in the sense that this ruling might support it) because this case does not concern human claims of authorship at all. It concerns undisputed solely-AI creation.

> you can't copyright something that your AI generated

Seems like a loophole, if I generate synthetic data with a model trained on copyrighted works, the synthetic data is copyright free? So I can later train models on it?

That's a legal implication. I'm asking what is it a practical implication. Why would an AI want to copyright their work?

Not open source...... public domain. There is a big difference.

When I have a LLM that spits out code identical to copyrighted code can I then use it legally?

Otherwise I would need to check the output of every LLM for copyright infringement

Code that the LLM reproduced without modification from it's ripped off "training set." I literally have no idea what kind of deranged person does not notice this let alone believes that they should profit from it.