AI generated content.
lorddayradon
Posts: 471
Just curious as to what the viewpoint or rules are with regards to submitted products that use AI generated content as textures. Some AI platforms deem the content created by an individual with their AI, to be works of their own, and are free to do with it what they will. Does/would Daz consider that an original texture? As AI learning in this genre, is heavily based on submitting millions of images, photo's, that it analyzes to learn that this is a hand, this is a car, this is a door, etc. Some I'm most certainly sure are others creations. Though when it comes to the AI actually making said, item/image, it only uses it as a reference, and has enough random variables with huge ranges, that the resulting image, may look similar to anothers content, but is not a 1:1 copy. This is easily seen when using the "make variation" feature.
Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
See https://www.daz3d.com/forums/discussion/441452/ai-is-going-to-be-our-biggest-game-changer/p1
Does that mention the actual official rules for submitting content as it relates or just the plebs debating back and forth about their and daz's interrpretation of AI in general? I don't need to go down a rabbit hole atm.
If your making textures. like seamless wallpaper or whatnot, with AI, to be incorporated with a product or render there's nothing wrong with that. The issue is when people claim a piece of AI imagery (why call it art) as something it's not. Falsifying or misleading it's origin or creation, that's a lying and basically unethical.
As far as sampling goes..well there's going to be a reckoning in the courts sooner than later on that.
But not all AI systems are created equal, but the principle is basically the same. Create a bunch of random noise and start filling in the splotches of noise with images based on the keywords via a neural network. So yeah, it's pasting pieces of other's work to fill those holes, it may pull from dozens and slap them on top of one another creating so much confusion it's impossible to tell but it's there.
However copyright law in the US protects even a portion of an image, not just the whole thing. It also protects "likeness" in some cases. So yeah draw a picture of Stallone he can techncially sue you unless you can prove it was parody or some other fair use. This is why you will see some sites be very strict regarding "Fan Art" ...but yeah others whatevers.
@charles I am not sure, but you might be wrong regarding the textures or wallpapers.
What I read so far about the TOS:
The AI companies declare all work created with their tools to be their own (the AI company is the copyright holder). But they allow you to use their work as you like - at least for now.
DAZ, as far as I know, requires the PA (you) to be the copyright holder. So, according to my understanding it is not allowed to make DAZ products with AI generated components.
But it would be helpful to get a message from DAZ to be sure.
EDIT:
It might be different with the "Stable Diffusion" open source solutions. If I run those on my local PC there is no coyright claim. But I don't know if an image created with an open source solution requires to be open source as well. In that case the question would be, if DAZ allows a PA to use open source components in a product.
Here's my license for anything I generate, Snip from TOS
4. Copyright and Trademark
In this section, Paid Member shall refer to a Customer who has subscribed to the latest Phase 3 payment plan, which became available as of 5/6/22.
Rights you give to Midjourney
By using the Services, you grant to Midjourney, its successors, and assigns a perpetual, worldwide, non-exclusive, sublicensable no-charge, royalty-free, irrevocable copyright license to reproduce, prepare Derivative Works of, publicly display, publicly perform, sublicense, and distribute text, and image prompts you input into the Services, or Assets produced by the service at your direction. This license survives termination of this Agreement by any party, for any reason.
Your Rights
Subject to the above license, you own all Assets you create with the Services. This does not apply if you fall under the exceptions below.
Please note: Midjourney is an open community which allows others to use and remix your images and prompts whenever they are posted in a public setting. By default, your images are publically viewable and remixable. As described above, you grant Midjourney a license to allow this. If you purchase a private plan, you may bypass some of these public sharing defaults.
Exception 1: Non-Paid Members License Terms
If you are not a Paid Member, Midjourney grants you a license to the Assets under the Creative Commons Noncommercial 4.0 Attribution International License (the “Asset License”).
The full text is accessible as of the Effective Date here: https://creativecommons.org/licenses/by-nc/4.0/legalcode.
Exception 2: Corporate-User License Terms
If you are an employee or owner of a company with more than $1,000,000 USD a year in gross revenue, and you are using the Services to benefit your Employer or company you must purchase a corporate membership plan to use the Services or copy the Assets for your company.
EDIT: I suspect there will be confusion in this. I'm no pro in copyright law... but as I comprehend it, there seems to be a difference in giving someone the right (license) to copy and reproduce vs being the owner of the original work which would mean that I retain ownership and all (master?) copyrights to it so to speak. Least that's how it reads to me.
Good, that you have clarified Midjourney licences.
Even while using Stable Diffusion on you own computer, it relies on training from images on the internet.
I am very tempted to try Stable Diffusion for my images, but not for selling anything created with it.
I have already seen many AI image based assets on the other sites for sale,
so it looks like people has found a way to monetize them.
I am also no laywer in any means.
But this one is interesting:
"Subject to the above license, you own all Assets you create with the Services."
I have read different statements but not sure if from Midjurney.
Anyway, I postet a support ticket to the sales department asking about this topic.
In general there is a lot of talk about whether those AI images break copyright of others in general - regardless of what their TOS say.
I for myself decided for now to not use any AI parts in commercial work until the legal situation is more clear.
Here's my thoughts about AI art and the concept of theft. In short, all art is stolen really.
There are so many facets to this topic, both for and against, it will be a long debate for a long time. Camps have and will continue to form around them much like many current long time debates. How long has the PC vs. Console debate been going on now?
That said, normal structure would dictate that a synopsis of my POV and then supporting reasoning, how ever I feel since this is such a fresh topic (and volatile), it's best to form it the other way around.
Art theft, being one of the main concerns, and rightly so, I feel is the strongest CON against AI art. Whatever I've done, I know, that if someone came along and copied my technique, or work, and then sold it as their own successfully (not gonna touch on effective marketing techniques) I would be angry. Frankly I feel it would also be depressing, that all my hard, creative, novel work, was going into someone else's pocket as their profit, or worse still, on someone's wall for free.
That said, there is a certain amount of free publicity and traffic or interest generated in your work, when someone has a background on their computer of your work but that's a minor point and could digress a whole page to.
My main feeling of why AI art theft is not really as big an issue as it's currently being made out to be as the big caveat is as follows in a common scenario.
(hypothetically) I'm a budding artist, and want to learn and improve my skill in my chosen field or style. As part of my learning process, and practice regime, I frequently head to various art galleries or exhibits and practice drawing in various styles and various content or subjects. As the months and years pass, I improve, to the point that I can now paint or draw, various styles and content exceptionally well. My work also enters a stage where it is now actually being sold for more than a pittance.
So am I stealing the art that I am using to improve my skills? Are my works that are being sold considered art theft? I think not, unless I am trying to make 1 to 1 reproductions and claim it as mine. It really doesn't matter what it is. Ice carving or sand sculptures, sidewalk chalk drawings, or even carpentry. Most if not all, art skills are perfected through attempted reproduction, albeit either an actual subject or someone else's reproduction. If I wanted to learn to draw a hand in various poses and angles, I've got lots of real hands all around me that I can practice drawing. Conversely, If I'm interested in drawing lizards or more specifically Komodo Dragons, there's not that many about for me study and practice with here in Canada. So as circumstances would dictate, my reference material would not be an actual Komodo Dragon, but rather photo's, paintings, drawings, etc. of others, at least until I could fund a $7k++ trip for a week to go practice.
That later situation is pretty much how AI learning works with respect to image : text correlation. If it's ok for a person to use that technique to learn and improve, I don't see why it should be any different for a computer.
Now personally, I can't draw, paint, or sculpt very well, and feel I'm much to old now to practice enough to be any good at it before my time on this earth comes to an end. I do however (imho) possess a good creative eye. I can see many things in my head, and know that if I could just pull it out it would be great. In short, my creative concepts are extraordinary, but my implementation fails. AI art brings me closer (at least with respect to images) to getting that image out of my head and into the real world.
The second major Con against it, is as mentioned is that it takes little skill, and thereby takes from the hand of an artist who has put years into perfecting their craft. Again, the raw nature of this concept if it was implicitly truthful without other considerations is a strong argument. I feel that again, this concept is unbalanced in it's portrayal. Let me explain. If I point to a plane that is in the middle of taking off, the front wheel off the ground, the rear wheels still on the ground, and proclaim "See that plane can't fly" simply because not all the wheels are off the ground, and hence it's not actually flying, I'm kind of distorting the truth and misrepresenting it. It's the same here with this Con.
One day you're the new and popular artist, raking in the money, and then along comes someone else. They could even be doing similar works (though distinctly different) even copying your methods. Think of that paint bucket suspended on a rope, spinning (orbiting?) above a canvas on the floor, and paint falls out holes to create spiral type patterns. If I use that method but say, add sparkles, or UV reactive paints, or even beetles mixed in with the paint that can fall through the holes, am I really taking from that artists hand simply because I used their method or tool, in a slightly different way, and now everyone is buying my art and not theirs? My impression here is that it is a very competitive field. You may develop a new and novel way to make art, and yes someone will come along and build upon and modify that and make money or notoriety from it. Their not taking from you. They just thought of it before you.
The skill debate also brings us to prompt craft, and yes that takes skill. I fell down a rabbit hole one night with Midjourney and ended up with the concept " An eagle eating a rabbit, that is eating broccoli" (don't even ask). I then got a series of results of an an eagle on the ground with broccoli in it's talons, or thereabouts. It takes a good vocabulary, strong grammar and structure to manipulate words, finding synonyms, or even variations of a concept through word variation to achieve your wanted results. Of course this is influenced by what the AI has been trained. A rather mundane example would be "Fog at ground level" vs. "mist" Sometimes "mist" used in conjunction with "waterfall" just doesn't get it done. You have to get creative and use a "fog" based prompt to get the effect you're looking for. This is experience based and creative skill based. That is prompt craft. (edit) One look at at my MJ archives can clearly show just how many changes, drastic or slight, that I go through with prompt craft before I get the image that is close enough to what I see in my head. Sometimes there is more than 30 of the same image with only a few changes. It's no less skillful imo, but it is faster, compared to trying to draw, paint, or sculpt, 30 works.
Copyright for now is the big AI killer.
The problem with AI generated content is it is really merging content from stock sites off the internet, so in reality the AI images are made from something the AI provider never owned in the first place which is why AI content is is banned from some stock libraries I believe and the ones I am with have sent emails saying dont even bother.
If you sell an AI image to say a publisher and they use it and one day down the line someone proves to them that an element of the image is theirs and sues them or even just threatens it they will sue you to the value of many many multiples of what you might make as the AI provider will sidestep the whole show saying you own it.
In the past there have been instances of books printed and then recalled and pullped and reprinted and that cost will be recovered from you this has happened in the past and when you sell images to the publishing industry they will often require a release you have to sign saying this is all my own work etc etc for this reason. This is why the commercial use element of Daz makes using it worthwhile and why the editorial license is worthless.
Shutterstock has said it will compensate artists for having thier imagery used in AI works on thier platform. Now I dont know if anyone here has ever used shutterstock but you get paid a pityfully small amount of money for each download. a few years back a photo of mine was used on a book cover in the UK and I got 15 cents and despite the use of the image exceeding the usage limit for the 15 cents the image was used in a major advertising campaign in the Guardian and Telegraph newspapers as well on the book itself. Shutterstock proved to be absolutely useless in getting the appropriate fee so I deleted my shutterstock account so as not to have anymore highvalue sales happen i would not get paid for. So as you might guess shutterstock will just put everyone over a barrel and keep on making money but you wont.
However if your creating an image for yourself there is no harm but if you want to sell it and even sharing might get problemmatic is another story. I am bitter though LOL sigh
this was the book
The question is, what does 'owning' mean?
One can buy a car and 'own' it, but one can't sell a 3D model of that car.
I get what you're saying and that truly sucks. Nice book cover. Did you have an actual feather present when you were composing it, or did you use someone else's images for refererence? If so, did you not steal from them? Same as art styles. If I paint in the style of Minimalist, I didn't create the style or the elements that make it. I'm essentially stealing the original artist's style elements. Can a style be copyright? Depends on who you ask. Giorgio Armani would say yes probably, whereas a carpenter whom has created a new style of staining wood or aging it might not.
I photographed the feather which was placed on an old writng slope i got in a jumble sale the designer of the cover tinted it.
OK so I found and selected the feather placed it on the slope and I lit it with 2 Elinchrome strobes and photographed it with a full frame DSLR and the postprocessed in Lightroom and Photoshop to remove dust and noise and thats before you have to go through the submission process to a website.
So if I see someone typing a phrase into an AI and getting a picture back and declaring themselves an artist frankly I think they are deluding themselves. but like I said I am bitter.
AI generated art is always derivative.
THe End User License Agreement, say it all.
When you pay for a 3D model, you are paying for the LICENSE TO USE.
The end user NEVER owns the model. THe COPYRIGHT OWNER continues to own the model which they LICENSE to others.
Entirely different from buying a real car out in the world.
Cars are even starting to come with EULA that restrict what you can repair yourself, it's been a thing with luxury vehicles, sports cars, some trucks and farm equipment for a while... one of the most egregious examples being John Deer equipment which will brick the machine if there is an unauthorized attempt at repair.
Look for far more of that as electric cars become predominant and the vehicles will legally be treated more like "devices" and less like traditional property.
Self driving cars will only hasten that departure from one owning their vehicle to simply buying the privilege of possessing something they don't actually have any rights to, like most electronic devices.
Dont get me started on BMW's stupid subscription service for heated seats and deluxe features... ugh.
Well at least air isn't a subscription yet...
(checks internet to be sure)
Phew... well... for now at least.
Note: Not a rant against electric vehicles (Which we desperately need), but consumer rights... (which we also desperately need more of).
I don't think AI is your art. The kicker in the post is what they have the right to do with your creation.
Rights you give to Midjourney
Using the Services, you grant Midjourney and its successors and assign a perpetual, worldwide, non-exclusive, sublicensable, no-charge, royalty-free, irrevocable copyright license to reproduce, prepare, and prepare Derivative Works of, publicly display, publicly perform, sublicense, and distribute text. Image prompts you to input into the Services or Assets produced by the service at your direction. This license survives termination of this Agreement by any party for any reason.
Please note: Midjourney is an open community that allows others to use and remix your images and prompts whenever they are posted in a public setting.
I don't Think I lIke that. AI grabs images from everyone on the net that uses that program. Then putting it together seem more like a collaboration on an artwork. I, in good consciousness, could not do that and claim it as my art. Copyright law is going to get real interesting, y’all
There will soon be no more“Artists”..only
“Art Directors”
What I meant was, if I take measurements of the car I own, and make a 3D model of my car, I still can't sell that model.
Rendo is allowing commercial license for AI-generated images in their freebie section and making a special gallery so AI users can spam their galleries into submission. I am disgusted.
I would also mention for environmentally-minded folk that there is a steep electrical cost for using the AI generators. I would have to find the link, but I've read that once the AI gets going, it will entail a 10-fold increase of cloud computing capacity to support the extra load of the AI generators and their content. And naturally, there is the hardware for all that, as well.
There are certain jurisdictions in the USA and South America (Bolivia) that try to make it illegal to capture rainwater, and some others that restrict ones ability to capture sun rays using solar panels without a kickback to the local utility companies. The absurd is already here. I know you're making some well appreciated humor, but the truth is only slightly hidden in the jest and potentially very scary.
Yup, this. I also had this thought with the rollout of the new 4-series graphics cards. The recommendation is now a 1,000 watt PSU...and there are stories of these cards melting adapters because of how hot they get with pulling as much power as they do. Where does this leave the environment?
The argument I always see made comes down to this:
Humans derive their style by looking at other works of art, so how is training an AI by looking at works of art any different? Aren't these both guilty of stealing to some degree?
The flaw with this argument is very simple. This argument is effectively equating the AI with a human. To make this argument, you must also be able to argue that AI is human, which is an argument you cannot win, at least in 2022. Humans will study art. They will think about what the art means. They will think about how specific aspects enhance the art. They will consider how they FEEL when viewing the art. Ect, ect. The AI does not do this. The AI uses pure computation to create the derivative works its creates. The AI is a machine. The AI does not have the rights that a human does, because it is not human. Thus the laws governing AI are going to be different than the laws that govern humans. Thus the argument that training AI is just like a human training has no merit. The AI does not enjoy the same benefit that a human does.
Right now the laws are not in place to keep up with technology. It always takes a few years for the law to catch up. So right now it may be totally fine to use AI generated works for a variety of uses. But I warn everyone not to rely on AI too much for their commercial products. The laws could very well change in a few years (or sooner as this gains momentum). So you may be fine for now, but do not expect that to last.
I believe that laws will be enacted based on how the AI was trained, because that is where most of the problems lie. Nearly all of these AI generators have been trained on some form of copyrighted work. After all, how does the AI know what Micky Mouse looks like? It was trained on images of Micky Mouse, that's why. The same goes for mimicking any artist. If you make the AI prompt that says "in the style of _____", obviously the AI was trained specifically on works made by that person. And this is different from a human studying art, as I already described above, the AI is not allowed the same freedom a human has in this regard because the AI is not a human.
And if you remove the copyrighted images from the AI, it loses its effectiveness immediately. You can no longer write a prompt asking for a Micky Mouse drawn in ____ artist style anymore because the AI doesn't know what these names are.
So I believe it is only a matter of time before countries start to enact laws to address AI. It will only take a few countries making laws on AI to create a ripple effect around the world.
I believe AI can be very cool. I certainly don't deny that. But there have to be rules in place regarding its use.
I half joke mostly, it's absurd the level of shenanigans currently engaged in by corporations and utilities... another disturbing example was back around 2015 when small city down south couldn't get the telecoms to install high speed internet, so the city did it themselves and immediately got sued by the telecoms... so basically, "we aren't going to do it, but don't even think of doing it yourself, just in case we ever decide to"... in that case the telcos eventually lost a court battle, but numerous small towns who've tried the same have been prevented from doing so due the the prohibitive cost of such litigation... it's absolutely dystopic the erosion of consumer rights that have occurred over the last two decades.
Every behaviour is always derivative as it's always a response to the current environment, so that includes art responses to the environment, an environment which include other behaviours as well. Derivative is such a weak term, and needs much more exacting of a definition for product X, when people try to force that word do serve them as a generic catch-all of ownership.
Say you wanted to reproduce the Mona Lisa. Traditionally you would paint that by hand and skill. There would be no dobut unless you were the most extremly talented painter that your painting by your hands is anywhere close to the original to cause any kind of confussion between which is the real one to which is yours. Now lets say you goto the louve gallery website, find the image of the mona lisa, click on the image, click copy. Then you open Photoshop and click paste, then save. Which of these 2 is closer to how AI creates it's works?
AI does not "create" art.....It "processes" existing artwork into something different. There is no creativity to it, it is a processing algorythim
agreed, but give us some slack being a bit loose with the language here.
I am going to gently disagree. I have three separate perspectives on A.I. in art.
The first is due to a familial connection. My sister worked for two major modern art museums. My own tastes are more Renaissance and Dutch Masters but I appreciate Modern Art as I have become better educated about it. I remember visiting her once so she took me on a tour of the musuem where they had a mockup of an late 50's/ early 60's house slashed in half filled with iconic artists of the period. I was startled by the Rothko (a painting in two colors that was so incredibly resonant and powerful) and a Jackson Pollock (a painting that was a human made fractal!); these are paintings that must be seen closely in person to be appreciated. The second is that I know a few artists including a noted photographer, an artist who (among other things) designs the video imagery that tops a skyscraper in my area called SalesForce Tower (https://sf-towertopart.com), my brother is a family friend of one of the "Names" in art (https://walkerart.org/collections/artists/kara-walker), and an old friend is an actual AI artist (he writes programs that take uses photos he has taken and creates generative art from the photos. Finally, the third thing that shapes my persective is seeing artists who have creative and beautiful work using AI. I am greatly enjoying the artwork of Jepe who is an artist here; his AI work is unconnected to 3d work but is sophisticated and ideosyncratic.
So, my issues with AI aren't about AI per se but 1) a tool is available and too many people are making hackneyed work with it (DeviantArt is rife with blurry, oversaturated Hot Topic Goth garbage created by AI), 2) too many people are using source material that isn't theirs which contributes to the generic quality of the output, and 3) people looking it as a "Make Art" button rather thana tool like Photoshop or FilterForge.
If you give the AI nothing but a blank canvas, it can do nothing. It needs existing art to process.
Real "Creativity" can take a blank canvass and make something from nothing.
Yes, I think that is my point. A paint brush on its own can't paint a painting; an artist is required to yield the brush. Daz Studio just sits on a hard disk until the artist decides to open it and create the scene or load a preloaded scene. A camera doesn'y take photos on its own. They are all just tools. One thing that should be clarified is that there are different levels of AI. The kind that is probaly of interest is the kind where you define the prompts that thell the AI how to compose the scene. When that is happening, that is when the creativity happens. I am suspicious but i hope that all of the hack work will be less common but that is the bulk of what we are getting now.