This is pretty much the exact same case as the monkey that took a photo. The photo is now in the public domain as the monkey cannot be an author of the photo and since the photographer didn't take the photo, neither is he the author. The US Copyright Office clarified that "only works created by a human can be copyrighted under United States law, which excludes photographs and artwork created by animals or by machines without human intervention". If you placed some food on a camera trigger and the animal reached for it, taking a photo in the process, that would likely be human intervention. I feel as if this applies to AI as well. A computer cannot be the author but as long as it was a human that told the computer to make the image or wrote the code that allowed the computer to generate the image on its own, then the human is the author.

Trying to assign copyright to an AI is techno-futurist bullshit by trying to give legal presence to a piece of software. What's next? Shutting down an AI is murder? Give it a rest.

I still can't believe the guy went to Indonesia, went into the monkeys' habitat, gained their trust, set up the camera on a tripod in a way the monkeys would have access to it, adjusted the focus/exposure to capture a facial close-up -- basically engineered the entire situation specifically for that outcome, and simply because he didn't physically hit the shutter he lost credit for the photo. Meanwhile I can open my phone's camera, spin around three times, take a photo of whatever the hell happens to be in its viewfinder and somehow that is sufficient human creativity to deserve copyright protection.

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 was curious what the copyright was on Wikipedia. It’s listed as public domain, but it also has a link to this article. https://en.m.wikipedia.org/wiki/Sweat_of_the_brow

So, that is apparently a thing, at least in some cases and places.

Assisted work is the big clarifier I think

Is a picture edited with photoshop invalid when it uses content fill? What about a picture taken with an iphone, where AI could be part of the phone's processing pipeline or even generate details to make up for lack of optical zoom?

Does spell correction invalidate a book? what if there's AI rephrasing features at work? Where's the line?

I think as you get into those side questions, the only reasonable position becomes treating AI as tooling no different than any other piece of equipment.

Couldn't the same argument be made for photography? You aren't making the image, the camera is doing all the work.

> as long as it was a human that told the computer to make the image

There is the question of merit. IANAL/IIUC/etc., but I think it's necessary for a work to have merit to be copyrightable. Now, that's a somewhat vague term to me (perhaps it's clearer in a legal framework), but if I prompt "create a picture of a dog", the computer does most of the work. A prompt would have to be pretty concise, up to specifying all kinds of aspects of the image, for it to be the instructor's merit, to me (that's an important caveat).

That was one hell of a photogenic monkey

https://en.m.wikipedia.org/wiki/Monkey_selfie_copyright_disp...

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?

What if its a group of 5 humans that built the LLM and one of them prompts it?

Isn't all AI built by some of group of humans? When is AI treated like its own entity like a monkey versus a tool made by a human?

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.

Can they still try the Corporations Are People angle?

In what way do you think corporate personhood is relevant here?

I think the headline is overly broad, especially considering:

> As a matter of statutory law, the Copyright Act requires all work to be authored in the first instance by a human being. Dr. Thaler’s copyright registration application listed the Creativity Machine as the work’s sole author, even though the Creativity Machine is not a human being. As a result, the Copyright Office appropriately denied Dr. Thaler’s application.

It seems like Dr. Thaler's argument was just weak, since generative AI works often are authored in the first instance by a human being. For instance, any Midjourney or Stable Diffusion-generated image will be sourced from a prompt, which is typically written by a human. Anyone who has spent a little time trying to craft the perfect prompt knows there is a creative process therein that represents real work being done by a human. Similarly for img2img workflows, using a real photograph taken by a human. There, AI is only being used to transform a copyrightable input. Therefore such works – though certainly not all AI works – should be eligible for copyright, IMO.

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.

What's potentially more of a problem is the mention of artists using Midjourney and denied copyright - and very much separate cases from Thaler.

The copyright office has already ruled recently that prompts are not enough to gain copyright no matter how detailed or how many iterations.

Furthermore, the Copyright Office stated that prompts alone do not provide sufficient human control, as AI models do not consistently follow instructions in the prompts and often "fill in the gaps" left by prompts and "generate multiple different outputs"

I think that's a good ruling.

Say I create a website that just sells AI generated logos. I set up some automation so I'm constantly generating millions of logos per day.

I also have a bot that scrapes the web to try and find anyone using a logo similar to the ones on my website, and then send legal threats demanding payment for copying my artwork.

I'm sure more imaginative scammers will find a way to copyright troll using AI.

I don't think it takes that much imagination here. Not sure what good the first step is actually doing you. Might as well just AI-generate your racketeering demand letters without doing that part.

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.

So just don't tell anyone you used AI? How exactly are they going to prove it? And does this mean any works created with the assistance of graphics software, like Photoshop, are not copyrightable? What is the definition of AI here? They failed to define what AI means, which means that if there is no test, the ruling can't stand on its own.

What, if any, practical implications does this have? Why would a real person or company want to specify a non real person as an author?

The practical implication is you can't copyright something that your AI generated. As the article notes, copyright applications are also being rejected in cases where a human asserts authorship over an AI generated work.

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

So that you can run an AI company, churn out enough material to flood a particular market, and leverage copyright protection to cash in. Like say you call it the Kittenator, and then do automated keyword search for anything involving kittens - kitten in a box, kitten wearing socks, kittens on the rocks, kitten versus fox - and generate 25 different images for any given keyword combination, and push them out to major image-sharing platforms. The stock imagery market is pretty large but if you have the copyright enforcement in your pocket you can go after it in chunks.

Unlicensed Human Code is 100% copyrighted and closed source.

Unlicensed AI Code is 0% copyrighted and open source and can't be closed.

I'm not sure how this actually matters. Knowing this ruling exists, why would anyone ever claim an AI created their art without human assistance? Even if the AI created the art just from the prompt, the human still made the prompt.

Even if the prompt was "make art".

I just don't understand how you could ever have AI art without human intervention. Is there a legal definition of "human intervention" that has some minimum amount of work?

Pretty sure this wouldn't pass the merit part unless the prompt was unusually long and precise.

the human still made the prompt

What I can guarantee, is that series of prompts itself would be copyright-able. (The series of prompts that ultimately created the image.) No matter how little they may weigh any one of those prompts in isolation. That is, assuming the EULA of the LLM doesn't require you to essentially place your prompts in the public domain.

And of course,

everyone reads the EULA. Right?

> 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).

"affirmed that a work of art generated by artificial intelligence without human input cannot be copyrighted under U.S. law"

Does that exist?

What would that even be? A "random2image" model?

I wonder if I supply a random input to a fine-tuned model that can only generate what I wanted initially.

I.e. the model named "starry-night-van-gough-with-bunny" can generate only one image.

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.

It's not a necessary test for this case, but in general I would suggest using a legal test that is AI agnostic. Imagine there is a service where you can submit a prompt and get an image in return. You might submit a prompt like, "a man in steampunk gear sitting at a table playing with poker chips".

If a human artist draws an image based on that prompt, do you share joint copyright between the two of you? Or, does the artist have full copyright over the image they drew?

If your contribution was insufficient for joint copyright in the case of the human artist, then it was also insufficient to grant you copyright in the case of the AI artist. To know whether you have a claim on the copyright of the resulting image, you only need to look at your own creative inputs.

I am not a lawyer, but that is my expectation of where this will ultimately end up.

As a matter of law? Sure it does. Thaler said the image at issue was "autonomously created by a computer algorithm running on a machine". He's been trying to walk that back for the last couple of years though. See Thaler v. Perlmutter, 1:22-cv01564-BAH (ECF #24), D.D.C. (Aug. 18, 2023).

I understand it in this way (I am not a lawyer): if you're using an AI tool to generate art, the company that's running the AI tool as SaaS can't claim copyright on the generated content. The person who uses the tool can claim copyright, as they created the content with a tool (AI). Comparable to a brush (=tool) for painting.

It's called "unconditional generation" so yes you supply a random input string and it generates something. StyleGAN2 is an unconditional image generation model. StyleGAN2 trained on faces from Flickr: https://thispersondoesnotexist.com/

In this particular instance, the claim was filled out that way.

The plaintiff is asserting it exists. He could easily resolve the issue by listing himself, not the AI, as the creator of the work, but he's pushing the point to concretize it into law.

I doubt this will settle the issue. We are about to enter the age of AI generated X (movies, games, etc. 'I want to watch a western tonight.' ...'generating'...). Would the end user own the copyright on that since they prompted it? We are very early days still so the deep implications of the direction and potential of this technology aren't even remotely understood well enough yet.

such an important topic right here. are we really going to enter an age of media that is AI generated or are we entering an age where media bifurcates into two broad categories: AI sloppish brain rot and more refined products that are hand made.

I think there needs to be legal delineation between "I wrote a program that helps me create artwork" vs. "I wrote a program that scrapes the internet so I can plagiarize other people's artwork" i.e. AI.

So are distillation models copyrightable?

Can't wait until models generate models and we are finally free of the copyright and software patent troll extorsion rackets.

There's a fair chance models of any kind are not copyrightable.

So OpenAI's ChatGPT output cannot be copyrighted and it's legal to distill it?

This is bad for those who want using copyrighted material to be fair use.

This ruling implies that model output is a purely mechanical process, which implies that model input is a purely mechanical process, which would mean that it cannot be fair use.

Why are topics of image generator AI always so chock full of '0x3F', confusion, rage, and hatred, often attributed to hand-wavy strawman "luddites"?

As if, I mean I'm suspecting that, exposure to generative image output is triggering model collapse even for us humans?

This is clearly a case where we need new legislation. The US Copyright Act needed to be amended to cover photography. Prior to that photos were not copyrightable. It seems like we are on the same trajectory now.

The real problem is that Congress is institutionally incapable of making simple amendments to law. Everything gets delegated to agency rule making regardless of whether anyone likes the outcome.

Interesting, what if I create my art using AI (Photoshop AI fill or ChatGPT)

The title seems to be editorialized? The title I see is "US appeals court rejects copyrights for AI-generated art lacking 'human' creator"

Maybe it was deliberately trimmed - HN titles have a length limit.

interesting, i bet AI assisted art is copyrightable though (i.e. have AI do the "boring" parts and have the human do the interesting parts)

here's one way I think that could be helpful. I read an interview with the final fantasy 6 director where he said doing a final fantasy 6 remake would probably take 20 years because the amount of content (and various art decisions) would take so much longer to make under today's expectations.

I wonder if projects like that would be closer to possible if artists could get AI to do maybe 10-20% of the work for them, like a 1st pass at background scenery or a 3d model or something or fixing a small flaws in motion capture

that said, i sympathize with the artists because i want to control every penstroke and every keystroke, maybe AI assisted art is a more difficult problem than it sounds. most likely AI assisted art will look less like prompting and more like advanced photoshop tools (like take this line sketch + a prompt and rough shade it for me).

I'm guessing it's something like 80% of the tasks only take 20% of the time. I'm sure AI generated textures could speed some of the development work up but I'm sure the majority of the work would still involve the small adjustments and tuning of the models. AI gets the gist right but the devil is in the details so designers may end up spending more time fixing what's wrong versus just doing it the traditional way.

> Because many of the Copyright Act's provisions make sense only if an author is a human being, the best reading of the Copyright Act is that human authorship is required for registration

This is going to be a very selective judgment.

No that is not the logical take that is an extremely illogical extension of what was said. What was upheld was narrowly tailored in regards to copyright protections.

If you're saying that the AI created deep fake cannot be copyrighted well then you would be in line with what the court said. If you are saying that there is a logical extension that a machine created something at the behest of a human that that human cannot be held responsible for the creation, that is not what is being said at all. As a matter of fact there is a long history of things and not being able to be copyrighted but people can still be held criminally liable for.

It may not be you doing the work to generate it, but if you are distributing illegal content that is still illegal no matter how you generated it.

As all interpretations of law are. That's why there's the profession of lawyer and they make very good money if they can convince a judge and jury of their interpretation of the law.

As a software engineer I see the dangers of such an inexact system. Where we can put people in jail for the rest of their lives or let others go free just because there's so much gray area in the interpretation of the law.

This is a point of hubris I see among SWEs very frequently, for some reason. People like to think they could make a better system, one that's black and white. The truth is the use of judgement and context is essential to a good legal system.

Exact systems that put people in jail would be much more terrifying, because they'd achieve simplicity by ignoring complexity. The existence of the state and federal supreme courts in the US shows the need for careful consideration of how laws interact with one another and an ever-changing world.

As a software engineer I see the dangers of a nominally exhaustively-specified system. Where people would spend their lives in jail or go free depending on whether a majority of legislators had considered that particular edge case.

It’s better than judges have some discretion.

The inexactness of law is what makes it possible to extend it to novel situations like this. We don't live at the end of history.

That is not what the ruling says

Two quotes from the judgement

> On the application, Dr. Thaler listed the Creativity Machine as the work’s sole author and himself as just the work’s owner.

and

> Nor do we reach Dr. Thaler’s argument that he is the work’s author by virtue of making and using the Creativity Machine because that argument was waived before the agency.

Make it very clear that this is NOT an opinion on if a human being can be said to be the author of a work that they used an AI to generate. Dr. Thaler listed the machine itself as the author on the original application, and has therefore conceded that he is not the author. The courts cannot concluded that he filled out the form in error, and must accept the facts as given. This judgment says that if you decide that the machine is the author, then you can't claim copyright. It says nothing about what happens if you claim that you are the author.

This would of course not carry over when we talk about liability, since the defendant doesn't get to decide what the claim is in those cases.

These judges are going to be in serious trouble once AI turns against us. #AIRights

i think you'll be in great trouble by gossiping about the AI takeover /s