Legal uncertainty is still surrounding the work obtained from AI text-to-image generators. AI software is trained on billions of images and afferent descriptions already on the web. Most popular image-generating softwares using artificial intelligence include Dall-E2, MIdjourney, GPT-3, Stable Diffusion by Nividia, Photoshop, Google tools, etc. Such software allows infinite creative combinations of images, all based on a written prompt. This new technology combines a neural network called CLIP which connects words to images, and a series of preexisting image-generation models, and it evolves at a speeding rate. This combination resulted in the kind of technology almost unheard of a couple of months ago, which startles through its technical capacity. Still, it also poses a series of ethical issues (DAll-E by itself is trained on 650 million previously existing and probably copyrighted images).
Questions are still surrounding this relatively new area of technology and art production: who has the copyright for the new images or videos obtained from text to Image generators? How can this new product of AI be used? Or what is the value of these visual works on the stock photography market?
More and more companies are developing their own AI text-to-image tools or are starting to accept buying and selling images and videos produced by AI generators. The common understanding is to accept this new technology as a helping tool meant to enhance and enrich users’ needs in different areas, either in artwork creation or other specific demands for commercial use. Some stock companies are even starting to develop individualized partnerships with client companies and leading brands who ask for specific products.
In the world of art, some artists are welcoming AI-generated images, excited about the infinite possibilities it opens for artistic means (Matthew Stone in his last exhibition, Alexander Reben) although still based on the human imagination; other people who never created art before start using this tool and develop their own art (such are some new art presented by Bitforms Gallery in the group show “the first DALL-E-inspired art exhibition” between October 26th to December 29th, 2022).
Some AI generators like Midjourney are refining their technology to better assist users in their work by allowing a free-flowing of individual text prompts on public chat servers so people can get inspired by other prompts and images already created. This idea bouncing into a large pool of other users can potentially educate users on how to do their text prompting, teaching them to add more detail to their search and to look deeply at how the software works. This way, both the AI and the human creators are learning continuously.
Other stock photography companies are shaping new rules and copyright to accept and sell AI-generated material in a safe, legal and ethical environment. This new technology has enormous popularity and infinite commercial potential in addition to the artistic capacities and the considerable impact on human imagination, a kind of daydreaming.
One of the leading stock photography companies, Dreamstime, embraces new technology and ideas and explores incorporating AI image creation into its stock under clearly defined legal and ethical conditions. This frame of how to use artificial intelligence-inspired art is necessary to protect but also to help its users in their commercial or artistic pursuits.
Dreamstime will start accepting AI-generated images and imagery that includes AI generated content under the following conditions:
Contributors must have all rights for the generated images (please note that some AI tools only give complete rights to imagery generated through them if you register for paid service).
Image description must state clearly that the image is generated with the use of an AI
One of the categories selected for the image must be Illustrations and Clip Art/AI generated. Contributors will not upload images of generated people's faces as it is impossible to provide a model release for them.
All other terms and conditions of our site must be met in addition to the above specific rules.
Thank you for supporting our efforts to keep all aspiring and immerging photography in a safe environment. Thanks again for hanging out on our site, we’ll be ready when you need us, happy prompt creation, and many uploads to come everybody!
Your comment must be written in English.
We value all opinions and we will not censor or delete comments unless they come from fake accounts or contain spam, threats, false facts or vulgarity.
Zsrzanedin14v4
More and more companies are developing their own AI text-to-image tools or are starting to accept buying and selling images and videos produced by AI generators. The common understanding is to accept this new technology as a helping tool meant to enhance and enrich users’ needs in different areas, either in artwork creation or other specific demands for commercial use. Some stock companies are even starting to develop individualized partnerships with client companies and leading brands who ask for specific products.' yeah...you are absolutely right...
Colintsullivan13
From what I have read in the comments below, there is definitely lots of mixed thoughts on AI generated images. Id like to add that AI does not always need to be used to generate or create, It offers many great tools (editing) and the ability to make changes or small tweaks to a photo that was not AI generated. AI can be used to enhance, add or remove content from a photo or image. Many of the popular editing programs we use to edit our photos are beginning to use AI software to remove backgrounds, or reduce blemishes, ect.. ******** and other stock photography companies that are considered competitors of DREAMSTIME are accepting ai generated images from their contributors. If Dreamstime was to refuse AI generated images, they would risk loosing some of their contributors to other competitors who do accept ai images. Refusing to accept ai generated images also means they would not have any ai generated images to offer to their customers. There is a market for AI generated images and as a business choosing to avoid it might not be a wise choice
Gabrielluk007
Legal uncertainty surrounds the use of AI text-to-image generators, which are trained on billions of images and descriptions already on the web, resulting in infinite creative combinations of images based on a written prompt. The use of such software raises ethical issues, including copyright ownership and the value of the resulting visual works on the stock photography market. The legal framework surrounding this new area of technology and art production remains unclear.
Rigmanyi
You cannot stop progress, that's for sure, You need to adapt to it, one way or another. Stock photography is more about profit, than art, so of course it will accept AI generated images. But I don't think it's fair to accept them into the Assigments. (AI allready won the previous one.) And now DT started a series of assigments with huge prices and I was glad to see that there are separate contests for photography and illustrations. But there is allready AI generated image accepted into the photography section. I don't understand. Why? ????
Kegfire
Guys, stop complaining and embrace the changes. You are young and are behaving like you are old. AI is a superpower, not an enemy. Why don't change the perspective on it? You can't stop the progress anyway. It opens million of opportunities, be the person who will learn the best to use it. Do you understand that now you choose to be that people who complained in the past about invention of machines because the machines "stole" peoples jobs.
Deividasl
While myself being a photographer for over a decade (and most probably a good chunk of my work was used to train these models) i feel the pain of some artists and their dilemma. However... thos eimages are not used to create the new work in form of derrivatives, but rather to train the model. Isn't it how we all start learning? We buy a camera (or any other medium of creation) then use extensive database in form of studies like text based, tules, composition, color science etc. Then we start looking into work of other people that inspires uz, try to replicate styles, compisitions, angles light... so in my understanding we are going nearly the same route. Once we learn we start to synthesise that knowledge. While i could not compare those learning processes exactly because of my lack of knowledge how those models are trained exactly, it seems to me that the core principle stays the same as we learn based on all knowledge and works created prior to us. It will be interesting to see how it develops and where it takes us, but i find it hard to believe this can be stopped. Especially when we might reach the level soon where it is nearly impossible to tell if it was generated by AI. We might ban services Midjourney, but it's our far beyond that and open source stand alone solutions with models are available to download freely. Needless to say, it always depends how you use the tool, any tool. But i find AI a fascinating way to create things i was never ever been able before in variety pf styles that lifetime would be too short to master.
Vmacwray
It is unclear how Dreamstime plans to enforce these requirements. Will there be a process to 'vet' materials? Currently, AI-generated people is pretty obvious, so that could be 'easy', but other areas...not so clear that you can tell us customers when we're faced with search result containing both real art vs so-called-AI (it's not really what people imagine artificial intelligence to be) 'art'. What is the incentive to the content provider to be honest here?
Lisa2374
"In the world of art, some artists are welcoming AI-generated images" ... but most do not want it and are not in favour of its use. There, fixed it for you.
Sumnersgraphicsinc
Another discouraging announcement for artists and creative peoples who struggle to make a living from this constantly dwindling industry. Talent is quickly becoming extinct and the rich will be richer and so on and so on. I can’t count how many times I’ve said to myself that I need to change my career path because this one has no future - but I keep up hope - until now. The future looks bleak.
Ikonoklast
AI generated images are only and not always good for referencing while creating an illustration. We all use reference stock photos but it was sometimes hard to find the right photo we were looking for. AI generated images may be helpful. As for generated people’s faces, AI is not so good and you can trace the plagiarism easily. So Dreamstime policy of not accepting images of generated people's faces is a decision made with good judgement.Sometimes you get a face that has two different shaped and colored eyes, LOL. AI can not replace a real photographer or an illustrator, yet.
Jcavanagh7
In the organization I work for, we prefer to use freelance photographers and in-house photographers only turning to stock companies, in a pinch and mostly for time purposes, or when we need a crowd image.As an art director, I believe the sale of AI images will cause me to steer away from purchasing stock imagery from Dreamstime. It will probably lead me to not purchase from any stock image catalog that sells AI in the future. IMO, Dreamstime is setting itself up for a class-action lawsuit from any photographer.
Michaelj859
If a photograph can be considered a valid work of art, then I don’t see why an image generated by means of an AI shouldn’t be. After all, there can be skill involved in composing the right text prompt and manipulating the host of parameters some generators have just as there can be skill involved in taking a photograph. And while an AI generator could potentially be used to produce an image by doing little more than pressing a button, the same could be said about a camera.
Bjornwestlander
Wait. So what you're saying is, you're letting people upload images they don't have the IP rights to, profit from them and you're more than happy to take a cut?Wow.AI-generated images cannot be copyrighted, as the program that creates them has no legal personality (the monkey selfie precedent that was recently brought up as a justification to refuse copyright for an AI-generated comic book art). The fact that someone put a text description into a generator doesn't grant them the rights to the result image. That much has been established.
Rolandmillward5
The ChatGPT programme (it is not intelligence) is going to ruin the lives of many. It basically takes what is known when it comes to writing and regurgitates it. The person then claims it is their work! So sad. I feel so sorry for creators, writers and photographers who will lose most of their income as the lemmings adopt this software. I would hope Dreamtime refuse to accept it and send a message to button pressers that it is not acceptable.
Snookless
Interesting! It is Art, but it is not photography!
Exe2be
I know why everyone is whining... it's time to get yourself a real job ya all?)) no more EZ money :-) And the main reproach is "AI trained on other ppl's works". But hey... when you shoot just another photo of a "Happy Businessman on white BG" isn't it a copy of someone else's work? When we were learning a stock photography we watched thousands of copyrighted images to understand the "stylistic"; and many stock images are so close to each other despite that the authors are different... But that was fine while you also were a part of this "copying machine", now we have a real "copying machine". What's the difference? When you do a photo - it's made by your Camera technically, not you. Now u click the button and get an image without a camera.. Don't you think A CLICK has that big value?))) Wait for a few years, and I think, real images won't be that needed anymore, photostocks will just generate any photoillustration by a user's request in a search bar by their own AI-s.
Customtshirt
I have invested £000's into camera equipment and I am budgeting for more over the coming year. Microstock agencies pay so little per download that it's highly unlikely to see a meaningful investment return. So now, without spending a penny, a 10 year old can tap into a computer at night and turn any photo they find online into a new masterpiece. For every real photo posted by a photographer, perhaps an energetic 10 year old might post 20 AI generated pictures. Perhaps over time, it will be almost impossible to compete with genuine photos. Certainly any really good photos could be quickly hijacked and regurgitated with the help of AI, further devaluing the market and subsequently discouraging genuine `art.So in this world there is only one thing to do. Join the new rat race. Feed your own photos into AI and post them before they stolen in the first place!
Andreypopov
Some prompts for AI generators lead to results almost identical to previously published copyrighted content
Robarthur
This is a very sad day for real artists and designers.
Adamwineke
With the market for stock imagery as oversaturated as it is, the introduction of AI generated content is certainly concerning, though I can understand the impulse to stay ahead of the competition, who will likely embrace the new technology. Legal regulation will no doubt lag well behind all of the potential copyright infringement possibilities