How does copyright work with output images?
If someone runs the model on their own hardware, do they "own" the images generated?
If 2 people generate the same image using the same prompt/seed, who "owns" the image?
In particular:
>“Even though you argue that there is some human creative input present in the work that is distinct from RAGHAV’s contribution, this human authorship cannot be distinguished or separated from the final work produced by the computer program,” the office stated.
The US does seem to be a bit of an outlier here. The above work was granted copyright in Canada and India.
That’s one case, not a ruling about all AI generated art. It won’t be the same for every image involving AI in some way. What if you use AI to fill in a portion of an image, as with Adobe’s content aware fill? What if you use a series of SD steps but with a human selecting outputs and feeding them back in as inputs to get something else the AI could not have come up with on its own? The copyright conversation is only just beginning.
>That’s one case, not a ruling about all AI generated art
"Because copyright law as codified in the 1976 Act requires human authorship, the Work cannot be registered."
The actual ruling (and a similar USPTO discussions) are about AI generated art and talk extensively about it in the broad case. The stance of these organizations is that AI generated art is not copyrightable. I don't disagree that the line is blurred when you discuss content aware fill, where the AI is working on a portion of it, but the current use of SD, even img2img and multiple prompts, etc., quite clearly falls outside of human authorship as recognized by the US Copyright and Patent offices.
Might this change in the future? Possibly. But as it stands today, I would not make any plans that assume you can secure the copyright (in the US) to anything made with SD.
Edit: Going through and noting that I'm not a lawyer and this isn't legal advice, don't listen to some random on the internet for legal advice, get a lawyer if you need it.
> quite clearly falls outside of human authorship as recognized by the US Copyright and Patent offices.
I think these are answering a slightly different question, as they are asking if the AI itself can hold the copyright on the output. A bit like if someone tried to copyright an image and assign “Photoshop” as the author.
The question above is maybe closer to asking if the person using an ML model can get copyright on the output, in that case there is a person trying to own the copyright, so I suspect it would not be rejected so easily.
>as they are asking if the AI itself can hold the copyright on the output.
Who is?
The original question I replied to:
>If someone runs the model on their own hardware, do they "own" the images generated?
This seems to be straightforward - Thaler tried to receive the copyright for the artwork generated by his Creativity Machine. He was denied, because the copyright office does not believe that a neural network generated image has human authorship.
From the Copyright office paper: "he [Thaler] was “seeking to register this computer-generated work
as a work-for-hire to the owner of the Creativity Machine.”"
>A bit like if someone tried to copyright an image and assign “Photoshop” as the author.
This is also clearly outside of the scope of copyrightable work per the reasoning given by the copyright office.
Both questions are thoroughly answered at this moment unless Thaler wins his appeal.
Edit: Going through and noting that I'm not a lawyer and this isn't legal advice, don't listen to some random on the internet for legal advice, get a lawyer if you need it.
> > as they are asking if the AI itself can hold the copyright on the output.
> Who is?
Thaler is. I’ve only read the intro sections of the documents you linked to, so I may have missed something more fundamental later, but the key points seem to be:
> The author of the Work was identified as the “Creativity Machine,” ... the Work “was autonomously created by a computer algorithm running on a machine”
and:
> Thaler must either provide evidence that the Work is the product of human authorship or convince the Office to depart from a century of copyright jurisprudence. He has done neither.
So in this case they are asking if the AI can be the author.
Whereas the question in this thread was:
> If someone runs the model on their own hardware, do they "own" the images generated?
In that case, a human is providing a prompt to the model (providing creative input), and asking if they themselves count as the author (a human rather than a neural net), so it seems like a significantly different case.
Thaler specifically asks for himself to be given the copyright assignment in the filing, claiming that the AI is essentially creating it in a work-for-hire. He does not ask for the Creativity Machine to be assigned the copyright.
>In that case, a human is providing a prompt to the model (providing creative input), and asking if they themselves count as the author (a human rather than a neural net), so it seems like a significantly different case.
I don't know that I specifically agree with this, but this is probably due to me having read additional articles on similar filings, including one where someone took a photograph, applied a style transfer AI to it, and then tried to copyright the resulting image, and was denied, because the copyright office found that there was not evidence that the work was a product of human authorship.
Andres Guadamuz (a lawyer specializing in IP law, senior lecturer at Sussex university, and a proponent of AI generated work being copyrightable) discusses a lot of this in https://www.technollama.co.uk/dall%c2%b7e-goes-commercial-bu... - but the most relevant part to this discussion is "For the most part, the legal consensus appears to be that the images do not have any copyright whatsoever, and that they’re all in the public domain."
The user experience for DALL-E, StableDiffusion, Midjourney, etc. are all essentially the same - craft a prompt, fine-tune it, get artwork out, so his discussion should be broadly applicable to all of these similar tools.
Thanks for the link, interesting reading. I can totally appreciate the angle that some generations may be too trivial to be worthy of protection.
I happen to be in the UK, and this happens to match my expectations, but it does strongly imply more regional variation than I’d have guessed:
> The situation may be different in the UK, where copyright law allows copyright on a computer-generated work, the author of which is the person who made the arrangements necessary for the work to be created. This, in my opinion, is the user, as we come up with the prompt and initiate the creation of the specific work. I think that there may be a good case to be made that I own the images I create in the UK.
This was a Style Transfer AI - it takes a source image and recreates it in the style of a painter.
In this case, the person both took the photo that the style was transferred to, and selected the style and a variety of variables. The US Copyright office still felt that his contribution was not distinguishable from the work that the AI did.
I'll note that this is very US specific - there are a lot of counter-examples of other countries allowing for the copyright of work like this, including the EU, UK, Canada, India, etc.
What if I create a fully automated image site where people can purchase images. All the images are generated by SD based on keywords I scrape from competitor websites. Would be quite easy to create a website like that.
Is it any different than Photoshop content aware fill? Or using a camera?
Nobody would ever think about Adobe or Nikon having copyright claims over your pictures. For me it's just a tool, the artistic part is providing a good description/base image, refining and choosing the best output.
Anyway, I'm not a lawyer and we probably live in different countries, so it'll be interesting to wait for the first lawsuit.
> Nobody would ever think about Adobe or Nikon having copyright claims over your pictures
But it is illegal to share pictures of Eiffel Tower, for example.
People do it, but they shouldn't.
If I put a picture of Eiffel Tower at night in a book or any other kind of commercial product, I have to pay to use it. Doesn't matter that it's there for my eyes to see it.
The question is: are the images generated by an hyper accelerated learning machine using copyrighted material without the author's consent legal?
I think they shouldn't be and the data included in the training should be free or licensed.
