In August of 2022, an artist whose name nobody recognized at the time, entered a Colorado State Fair fine arts competition. The artist, Jason M. Allen, won first place in the emerging artist division’s “digital arts/digitally-manipulated photography” category. Soon, everyone would know this artist’s name, but most people would not consider him to be a “real artist.”
After winning first place, Allen excitedly posted on his MidJourney topic Discord server on August 25th. He showed off his winning piece, titled “Théâtre D’opéra Spatial” (French for “Space Opera Theater”) with the first-place blue ribbon. The post was shared on X (known as Twitter at the time) and it quickly went viral.
The word was out: Allen used Midjourney, a popular AI Generative art program, to create his work of art. He did not paint it or digitally draw it, and instead used text prompts and an AI program instead, to create three images, one of which won the competition. X users let Allen know exactly how they felt. One user wrote, “This is the literal definition of ‘pressed a few buttons to make a digital art piece.” Another user said “This sucks for the exact same reason we don’t let robots participate in the Olympics.”
Allen did not necessarily hide the fact that he created his winning piece using an AI Generative program. Cal Duran, an artist and art teacher who was one of the judges for the State Fair competition, said that while Allen’s piece did include a mention of Midjourney, he did not realize that it was generated by AI when judging it. Duran sticks by his judgment.
Through this popular occurrence of Allen being damned for creating art using an AI Generative program, we begin to see the disconnect that exists between AI Generative art and the artist.
1
1
1
Technology Enters the Art World
This is not the first time a new technology has shaken the art world to its core. The art world was forever changed in 1826 with the emergence of the camera, and when Joseph Nicéphore Niépce took the world’s first known photograph. Like with any new technology, some artists were curious to test the new art medium, while others looked down upon it and saw it as a threat to the established world of traditional art. The camera could instantly capture exactly what the eye sees, something that a painting of a two-dimensional image could suddenly not quite compete with. Slowly, digital technologies like the camera ushered in a new form of art. Artists and audiences alike were forced to begin to reconsider their definition of “artist.” Should the camera user be seen as a valid “artist,” if all they do is click a button to capture an image?
In the mid-1960’s the term “computer art” came about. Artists who were savvy with both computers and art began to generate drawings with algorithms based on chance, that produced shapes and elements to achieve digital artworks. By the early 1980’s the term “digital art” was created by artist Harold Cohen, who worked with computer engineers to create a machine called AARON, which is seen as the earliest form of painting software. AARON was a robotic machine that was programmed to create drawings on large paper sheets placed on the floor. By the late 1990s, the digital art age greatly advanced with the emergence of new technologies and became popular and easily accessible due to the rise of the personal computer. Digital art started to be used for illustrations, websites, and even in commercial advertising. Almost anyone with a computer had easy access to be able to experiment with technology and create works of art that had never existed in the general public before. Society was once again faced with continuing the debate that the camera started – but this time, should the computer users be considered “artists,” or is the computer and software program doing all of the work?
2
2
3
Digital Art and New Media
Digital art is defined as any creative practice that uses digital technology as an essential part of the artistic process. Just like traditional fine art, digital art offers multiple mediums, from digital photography to computer graphics and pixel art. Most notably, digital art is placed under the larger category of “new media art,” which refers to a wide range of contemporary artworks that include new technology. New media art is intrinsically tied to digital art and encompasses all types of contemporary art created, edited, or broadcast by new media technology. Some examples of new media art include computer animation, video games, and robotics.
​
Thanks to the digital age, the definition of new media art was created. This means that it is acknowledged that as new technologies are invented and adopted into the art world, that new forms of art way beyond traditional painting will come about. However, there is one form of new media art that is bringing the familiar debate back into the art scene, and that is Artificial Intelligence (AI) Generative art.
4
4
5
AI Generative Art
The term Artificial Intelligence was coined by computer scientist John McCarthy long ago in 1955, to refer to “learning and thinking” done by machines. Today, this concept has rapidly evolved to encompass all types of creative processes that involve machine learning algorithms to generate different kinds of content, including images, text, videos, and audios. This, specifically, is AI Generative art. And there are many AI Generative programs, like DALL-E and MidJourney, that just like many of the new technologies and programs, are available to the public to access.
​
Right now, all AI programs are built with machine learning. They are given billions upon billions of text and image pairs. The programs are trained by being told to pair the images and the text together. It randomly pairs the pictures and words and is given its error margin, which it uses to get closer and closer to the correct answer. The machine does not actually “see” the images and it cannot cognitively be capable of making the pairs. Instead, it sees numerical values for each pixel (colors and shapes are 1s and 0s to the AI). Most AI programs have learned from The Large-scale Artificial Intelligence Open Network (LAION), a non-profit organization that provides datasets, tools and models to aid in machine learning. Both MidJourney and DALL-E were trained from LAION. Some AI programs also use a model of AI training called Deep Learning which is designed to mimic the neural networks of the human brain and how we make connections to understand things.
​
There are three common ways in which AI Generative programs work. First, there are AI algorithms that can produce images by taking information from big databases of human-created pictures, through a user’s text prompt. This is the most common. Another way Generative AI is used is by a user giving the algorithm a particular graphic image and then using text prompts to manipulate and edit the existing image. This feature is quickly becoming more popular and being built into existing editing programs. And the third one is a mix of the previous ones, in which a user gives the algorithm an existing image and through text prompts has it create multiple variations of the original image.
​
Remember Jason M. Allen’s winning art piece, using MidJourney. He used these generative techniques. He claims it took him around 80 hours and over 900 iterations to get to his final three pieces. He said that first, he said that in order to achieve a Victorian-style image with outer space themes, he experimented with the word prompt phrasing in MidJourney that led to the generation of images of women in frilly dresses and space helmets. He kept tweaking the word prompts to adjust the lighting and color in the composition, and after generating over 900 images, he settled on three. He then took the images to Photoshop to clean up some details, and then he ran the images through Gigapixel AI in order to improve the resolution so he could have the images printed on canvas.
​
Like new media before, with the rise of AI in the art world, we are again facing the familiar question: can an AI user be considered an “artist?” But, for the first time, compared to any new media before it, deeper, harder-to-answer questions are coming from the rise of AI generated art. There is a disconnect between the art and the artist like never seen before. Leading us to ask: who is the artist, the user or the AI algorithm itself?
3
6
3
1
Who is the Artist of AI Generative Art?
It might seem vapid to refer to Merriam Webster’s Dictionary for the definition of “artist,” but the definition helps to bring an interesting piece to the debate on whether a Generative AI program can be considered an artist or not. According to the dictionary, an “artist” is defined as a person who creates art (such as painting, sculpture, music, or writing) using conscious skill and creative imagination).
​
It is important to notice a few key phrases in the definition, especially “conscious skill” and “creative imagination.” AI is not conscious. While Generative AI programs are certainly impressive and may seem to be able to cognitively create pieces of art, it is all just a cumulation of their training of pairing photos and text that makes them appear as though they have any similar level of thinking close to a human’s ability. Remember that when AI programs are learning to match images to text descriptors, they do not “see” the images and they do not cognitively connect the two together. Instead, they randomly pair the text and images and based on their error margin can begin to make more accurate pairings. There is no “conscious skill” involved, simply randomization and an error report. Since AI is not conscious, it therefore cannot have “creative imagination.” The ability to be creative and have an imagination is what has been a large part of what separates humans from machines.
​
Referring back to Merriam Webster’s Dictionary’s definition, it could be easily assumed that the person using a Generative AI program to create a work of art is considered the artist of the piece. Not only does the user fulfill the human qualities of having “conscious skill” and “creative imagination,” the user also fulfills the “person” term in the definition. After all, the definition was formed because of the existence of human artists.
​
Creativity and imagination are important to the process of creating art, because it means some level of conscious thought has gone into the artwork. It is important that whatever is being created has some meaning, significance, and value to the artist. Of course, aesthetics are a large piece of creating an artwork, but it is undeniable that a human artist puts a lot of their own conscious thoughts, personality, and implications into the work, whether consciously or unintentionally.
​
Human inspiration and creativity often come from making connections, such as connections to the five human senses, various people and places, and other learned knowledge. Humans are inspired by a certain element, and then creatively produce another element that communicates their own connection to the original element. Artists are inspired to create by the artists that came before them, by the situations of their daily lives, by the pieces of knowledge they learn and emotions they feel, and so on and so forth. Inspiration is rarely from one thing, but from a connection of many elements important to an artist, both from their past experiences and current conditions. All of this is part of the human condition, an inner and outer environment that AI is totally isolated from.
​
Within creativity also comes the importance of originality. The idea of if something can be truly new and original, or if everything is built upon what has already existed and what knowledge an artist has, is still a difficult question with no clear cut answer. But it has been a long standing notion that originality is important to art. An artist cannot copy a piece of art or steal it and claim it as their own and call it creativity. And the concept of stolen art in relation to training AI programs has been a prevalent issue.
​
As mentioned, most Generative AI programs have been trained using LAION. LAION took almost 6 billion images of other artists' work from the internet, without anyone’s consent and without anyone’s ability to opt in or opt out of having their artwork be available to train the various AI programs. LAION was able to do much of this without anyone’s consent or interference, due to their non-profit status and claim that their work was all in the name of research. Therefore, any image produced by a Generative AI program is “copied” or “stolen” from an artist before its time, and the AI is unable to cognitively add its own knowledge, experience, or inspiration to the final piece, in order for it to be considered creative. Artists who have spent years perfecting their craft have had their art taken for someone else to quickly imitate using a Generative AI program.
​
An AI Generative program cannot match the level of conscious, creative input that a human creating a piece of art can. And at its most basic, an AI Generative program cannot do so autonomously, like a human can. The AI Generative program does not make art on its own accord - it always relies on a human to give it a prompt on what to create. Then, it creates an image based on its previous training, in which no cognitive abilities have been applied. There is no deeper meaning, significance, or importance to creating the image for the AI, like there would be for a human creating an image. The AI is simply creating because it was told to. The AI is also not creating anything original - it can only create based on the information and images it was trained with. The AI program is limited to what the humans who created it decided to train it with. AI programs will always be at the mercy of humans because not only did humans create it, but it requires humans to use it in order for it to be able to create any work of art. So, if humans create AI programs, anything created from an AI program is technically still human made.
​
This all leads us to believe that the human input outweighs the AI input, and therefore the work of art is the human’s. But, of course, it is not that simple, especially when it comes to matters of ownership.
9
9
9
10
8
9
AI Generative Art and Ownership
A suitable way to decide who the artist of a piece of work is, is to find out who “owns” it. Certainly, someone who uses a Generative AI program would not want someone else taking their text prompts or final images and claiming it as their own or using it commercially. This is the original issue with the images and art used for training the AI programs, so continuing the cycle of “stealing” would only cause more harm.
​
A simple way that modern artists can claim total ownership over their art and have peace of mind that it cannot be “stolen” in any way, is by having their work copyrighted. Interestingly, under copyright law, copyright protects original works of authorship from the exact moment that an original work is “fixed” in a tangible medium. However, sometimes artists go a step further to register their work under copyright law, to create a public record of their work in case they need to seek legal actions.
​
In order for any work of art to be protected under the Copyright Act, it must meet three requirements: 1. It must be an original work of authorship; 2. It must be fixed in a tangible medium; 3. It must have a minimal amount of creativity.
​
The Act clearly states that the AI program itself cannot “own” the artwork. It states, “The U.S. Copyright Office will register an original work of authorship, provided that the work was created by a human being.” Because of this, the U.S. Copyright Office currently does not allow anyone to copyright Generative AI artwork, because it is not totally humanmade. At this time, the office rules that AI generated art has no owner.
​
The AI programs themselves recognize this rule. For example, in looking at DALL-E’s terms of use, it assigns the right, title, and interest in and to the final image to the creator, provided the terms have been complied with.
​
This ruling was made clear to Jason M. Allen. Allen tried to copyright his Colorado state fair winning image that was created using MidJourney, and despite appealing to the U.S. Copyright Office three times, they made a final decision in September 2023 that it cannot be copyrighted. The office stated that his text prompts in MidJourney and visual modifications in Adobe Photoshop afterwards were not enough human touch to allow him to own the artwork. They believed there was too much execution of the technology towards the artwork, rather than Allen himself.
​
According to both the dictionary and the copyright rule, Jason M. Allen would be considered the artist of his AI-generated artwork. After all, he is human and therefore has the ability to have conscious skill and creative imagination. But according to copyright laws, Allen is not the artist and cannot own the artwork, because of the AI technology’s contribution to the work. The copyright law contradicts the definition, and states that the technology might be more of an artist than originally thought, despite the third rule being that the artwork needs to express the ability of creativity.
​
It should also be considered how much the artists of the original images are considered artists of an AI-generated image. The AI programs have used their artwork to learn, and therefore every artwork afterwards will be built off the building blocks of their work. But because of LAION’s nonprofit status, those artists relinquished their rights without having any control.
​
At this time, copyright law states that anyone who uses a Generative AI program to create a piece of art cannot claim ownership to the artwork, despite the authorship rules protecting the art as “theirs” the moment it is fixed in a tangible medium. This illustrates the great disconnect between the artist and the art that AI has created. But someone who uses a Generative AI program and considers it a tool, exhibits conscious skill of the technology, and brings an element of human creativity to the artwork would classically be considered the artist.
11
12
13
13
14
10
Is There More to Art Ownership than Copyright Laws?
So, U.S. copyright laws state that Jason M. Allen is somehow both the artist and owner of “Théâtre D’opéra Spatial” once it was fixed on a tangible medium, but also not the owner because the inclusion of AI technology was seen as too much of an interference, since he cannot register his piece with the office.
​
But as it has been made clear, especially with the new disconnect between the art and the artist due to AI, that definitions in the art world are not quite black and white. The terms “artist” and “owner” in regard to AI Generative art have become contradictory and confusing.
​
Artists create art every day with no intention of copyrighting their work. Whether for their career or for fun, an artist uses conscious skill and creative imagination to create a work of art. They autonomously draw on inspiration, previous experiences, and social values in order to create something of original authorship. It is created in their own style, for their own reason, and it is usually fixed in a tangible medium. So, why can't an artist use a Generative AI program and call it “theirs?”
​
If the artist uses a Generative AI program, creates it autonomously, and uses creativity and imagination with unique combinations of keywords, can the user not be considered the artist, copyright aside? It is the AI component that is creating this disconnect between the art and the artist, and if the user can be considered the true “artist” or not, because a work of art is being created both autonomously and non-autonomously. Both with conscious skill and artificial skill. Both with human creativity and computer coding. Both with human inspiration and devices trained with previously-created art.
​
Therefore, it is more difficult than ever to decide who is the artist of the artwork. When the camera came into the art world, it was debated if the user could be considered the “artist” with such an interference of technology. When graphic design entered the art world, it was debated if the user could be considered the “artist” with the inclusion of the computer and editing software. Over time, both discussions were resolved and today we regard many types of artists, such as photographers, videographers, graphic designers, and digital illustrators , to be true “artists,” even with the inclusion of the new technologies. And now, artificial intelligence involves a new level of technology interference that we have not had to face before.
What Other Artists Think
In order to get some other thoughts on AI Generative art, I interviewed three artists from three different disciplines: traditional art, camera art, and emerging technology and new media art. They were all asked the same questions, reflecting on what it means to be an artist in terms of creativity and inspiration, how technology plays a role in their art, their views on AI Generative art and an artist’s relationship with it, as well as various other questions. Because of their different artistic backgrounds, it is interesting to see how their opinions overlap or greatly differ from each other.
Julie Mikuzis: Traditional Artist
Julie Mikuzis has a wide array of artistic talents, from drawing, painting, and sculpting, to camera art and graphic design. However, her greatest passion is for the traditional, fine arts. Julie has taught at the high school level for over a decade, leading and inspiring students of all abilities, from first time painters to more seasoned practicers. Outside of the classroom, Julie creates drawing and painting both as a freelance artist, and for the love of creating for herself.
Connor O'Keefe: Filmmaker
Connor O’Keefe is a documentary filmmaker and editor. His short films have been played at major film festivals (Palm Springs International ShortFest, Nashville Film Festival, and Cleveland International Film Festival) and have been incorporated into a variety of other causes and events, including a TEDx Talk. In addition to filmmaking, Connor has been a part of various avenues of the film and media industry, including roles as an equipment manager, a stream technician, and a freelance editor. Connor is also a professor at DePaul University in the School of Cinematic Arts.
LeAnne Wagner: Emerging Tech Instructor
LeAnne Wagner is a full-time professor at DePaul University in the School of Design, in the Experience Design and Digital Communications and Media Arts masters programs. She teaches courses in interaction design, emerging technologies, human computer interaction, storytelling, digital fabrication and community collaboration. She is also the Director of DePaul’s SPARK (The Strategic Partnerships for the Advancement of Research and Knowledge) Center, which is focused on interdisciplinary and translational research. Leanne Wagner previously worked in the field of graphic design, and now works with students to develop their skills in AR, VR, and of course, AI.
Author's Take on Generative AI Art(ist)
I believe that the moment Joseph Nicéphore Niépce took the world’s first known photograph, artists and technology began to coexist. And with the introduction of new technology after new technology - from computers to editing software to robotics and, of course, to Artificial Intelligence- artists have had to continue to learn to adapt and find their place within this coexistence.
Like many, I have conflicting opinions of AI Generative art, because of the existence of both human skill and artificial intelligence. I surely do not believe the AI program itself is the artist of AI Generative art, but it undoubtedly plays a huge role, as does the art in which the program was trained on. However, my stance strongly is: the user of AI Generative art is an “artist.”
The user must bring their own technical skills of using an AI Generative art program, as well as their own creativity and inspiration in order to give the program a prompt. While the AI does a lot of the task work, it would never be done independently if the user, or artist, did not interact with it. Like many, at this time I see AI Generative art programs as an artistic tool.
Issues around AI Generative art programs are not black and white. We cannot discuss AI without discussing both ethical and legal issues. AI has come a long way, but it still has a long way to go. The same conversations were had with the introduction of the camera (now known as photography, videography, filmmaking) as well as computer and editing software (now known as graphic design, illustration).
The reason conversation after conversation about AI Generative art is happening is because it comes down to the fact that we have never seen a disconnect between the art and the artist like ever before, especially because of the dual aspects of human creativity paired with artificial intelligence. And we are bound to continue having these conversations as Artificial Intelligence is incorporated into many of our new technologies. For example, if I used an AI text generator, like Chat GBT, to write this, am I still the “author” of this piece?!
Please note: I did not use any AI technology to write this.
​
​​
​
​
​
​
1: Metz, R. (2022a, September 3). Ai won an art contest, and artists are furious | CNN business. https://www.cnn.com/2022/09/03/tech/ai-art-fair-winner-controversy/index.html
2: Kenny, S. (2021, December 13). How photography changed the Art World. Creative Pinellas. https://creativepinellas.org/magazine/how-photography-changed-the-art-world/
3: Artplacer. (2023, December 1). What is Digital Art?: A definition of a new medium and media. ArtPlacer. https://www.artplacer.com/what-is-digital-art-the-history-and-value-of-an-evolving-concept/
4: Bravic, L. (2023, November 23). Digital Art - definition, history & innovative art forms. Artland Magazine. https://magazine.artland.com/digital-art/#:~:text=Although%20the%20first%20digital%20art,and%2C%20especially%20the%20latter%2C%20envisioning
5: Berthonneau, A. (2022, October 5). New media art : What it is and how to create it! - heavym blog. HeavyM. https://www.heavym.net/new-media-art/
6: Hello Future Me. (2023, June 3). The Ai Art Apocalypse. YouTube. https://www.youtube.com/watch?v=9xJCzKdPyCo&t=4270s
7: Merriam-Webster. (n.d.). Artist definition & meaning. Merriam-Webster. https://www.merriam-webster.com/dictionary/artist#:~:text=%3A%20a%20person%20who%20creates%20art,conscious%20skill%20and%20creative%20imagination
8: Finkel, E. (2023, August 22). If Ai becomes conscious, how will we know?. Science. https://www.science.org/content/article/if-ai-becomes-conscious-how-will-we-know
9: Marr, B. (2023, October 5). The intersection of AI and human creativity: Can machines really be creative?. Forbes. https://www.forbes.com/sites/bernardmarr/2023/03/27/the-intersection-of-ai-and-human-creativity-can-machines-really-be-creative/?sh=208bcec73dbc
10: Malig, M. L. (2023, June 7). The controversy of AI Art Theft. Systemic Alternatives. https://systemicalternatives.org/2023/06/06/the-controversy-of-ai-art-theft/
11: Office, U. S. C. (n.d.). What visual and graphic artists should know about copyright. What Visual and Graphic Artists Should Know about Copyright | U.S. Copyright Office. https://www.copyright.gov/engage/visual-artists/
12: Copyrightable authorship: What can be registered. (n.d.). https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf
13: Mathur , A. (2023, September 8). Art-istic or art-ificial? ownership and copyright concerns in AI-generated artwork. Center for Art Law. https://itsartlaw.org/2022/11/21/artistic-or-artificial-ai/
14: Kenney, A. (2023, September 6). Jason Allen’s AI Art won the colorado fair - but now the Feds say it can’t get a copyright. Colorado Public Radio. https://www.cpr.org/2023/09/06/jason-allens-ai-art-won-colorado-fair-feds-deny-copyright-protection/
Sources
About the Author
Jill is a Chicago-based artist and designer, interested in emerging technology and its role in the art world.
This project was created for her June 2024 Thesis project. Jill is a graduate of DePaul University's School of Design, in the Digital Communication and Media Arts Master's program. Jill completed her undergrad degree in 2018 in Saint Xavier University's Communication program, with a minor in Graphic Design.
Acknowledgements
Thank you to Julie Mikuzis, Connor O'Keefe, and LeAnne Wagner for not only agreeing to be interviewed, but also providing tech help and content feedback. Thank you to Zammuto for allowing me to use your song "My Dog's Eyes" in my video. Thank you to Kathleen Miller for being my editor when I could no longer stand to read my own words. And thank you to my 2017 Mac laptop for slowly but surely sticking with me as I worked with giant multimedia files.