Artificial Intelligence versus/& Human Artists: AI as a Creative Collaborator in Art

April 16, 2024

Photo: Regen 1.0.2 by Colleen Hoffenbacker

By Atreya Mathur

“Recently, I’ve integrated artificial intelligence into my digital process to harmonize the cutting-edge technology with timeless human values, creating a blend that’s uniquely mindful and full of wonder.” ~ Colleen Hoffenbacker, Artist

In the realm of contemporary art, Artificial Intelligence (AI) has emerged as a groundbreaking force, revolutionizing the way the world conceives, creates, and appreciates artistic expression. From mesmerizing visualizations to thought-provoking installations, AI-generated artworks have captivated audiences worldwide, sparking discussions about the boundaries of human creativity and the role of technology in shaping our cultural landscape.

Under traditional copyright law, creative works are granted protection based on the principle of originality, which requires human authorship and a modicum of creativity.^[1] This protection extends to various forms of expression, including literary works, visual art, music, and film.^[2] However, AI-generated art blurs the lines of authorship, as it is often the result of collaborative efforts between human artists and machine algorithms. In collaborative works involving AI, questions arise regarding the respective contributions of humans and machines, complicating the determination of ownership and rights.^[3] While human artists may provide the initial concepts, guidance, and input, AI algorithms play a significant role in generating and shaping the final artwork. As a result, the traditional framework of copyright law may struggle to accommodate these hybrid creations, leaving legal scholars, artists, and policymakers grappling with unresolved issues of attribution and control.

In March 2024 at an exhibit in New York, Human-Assisted Art: Gallery Party, artwork from a wide range of artists and technologists, including interactive installations that explore the use of AI to expand human creativity were showcased. This included a “Poetry Camera” by Ryan Mather and Kēlín Carolyn Zhang and an “AI Photobooth” by Nikhil Thorat and Caroline McGuire, which serve as interesting examples of the scope of AI-collaborative art. The Poetry Camera is an inventive twist on traditional photography, generating poems instead of pictures with the help of AI. Utilizing a blend of Polaroid nostalgia and modern technology, the device allows users to select the type of poem it produces with a simple knob. Powered by GPT-4, Raspberry Pi, and a thermal printer, it captures an image and then prints a poem based on the scene. For example, when used during a rainy day in March at the New York art gallery, it captured the backdrop of artwork and two friends in blue and black skirts, smiling and laughing. The resulting print produced a corresponding poetic verse, which offered a novel way to capture and commemorate the moment with the use of AI technology.

Using the Poetry Camera at the Human-Assisted Art: Gallery Party, New York

But who would own the copyright to these poems generated by the Poetry Camera? Determining copyright ownership of such works could be complex and may depend on various factors, including the involvement of human creativity in the process. Since the poems are created by the Poetry Camera, it could be argued that the copyright belongs to whoever owns or controls the device. If a company or individual created and owns the Poetry Camera, they might hold the copyright to the poems it produces. Additionally, if the camera incorporates pre-existing poems or texts as part of its algorithm or database, the copyright to those texts may belong to their respective authors or copyright holders and may likely be being used without their consent to create these poems. Further, if users of the camera provide input or influence the content of the poems produced (e.g., by selecting themes or styles), they might potentially have some claim to copyright, depending on the extent of their creative input as well.

The AI Photobooth is also a fascinating blend of traditional photography and artificial intelligence. Similar to a conventional photobooth, users strike poses for a series of three photos. However, what sets this booth apart is its AI capabilities, which allow it to recreate the images using various prompts.

While using the photobooth, one particularly intriguing prompt included a reference to the renowned artist Greg Rutkowski, a digital artist known for his vivid, surreal style. Rutkowski’s name has become a staple among enthusiasts of AI-generated art, though he has expressed his disdain for AI-generated images both vocally and through his art. Despite his objections, prompts featuring his name continue to be widely used by AI generators including the photobooth, showcasing the issues that arise with AI tech including copyright infringement and ethical considerations while creating artwork based on another person’s work without their consent.

As the world delves deeper into the intersection of AI and art, it becomes clear that navigating copyright in the age of machine creativity requires nuanced considerations and innovative solutions. This article will explore the evolving landscape of AI-generated art, examining the challenges and opportunities it presents for creators and the broader artistic community, focusing on the collaborative role AI has the potential to play. The article also includes an interview with artist Colleen Hoffenbacker on her artwork, the artistic process in collaborating with AI to create art and registering her work with the US Copyright Office (USCO).

Decisions from the Copyright Office regarding Authorship of AI-generated Work

In a pivotal decision by the USCO, the registration process for “Zarya of the Dawn,” a graphic novel incorporating AI in its image creation, sheds light on the intricate intersection of AI-assisted art and copyright law.^[4] Authored by artist Kris Kashtanova, the graphic novel utilized the Midjourney AI platform to generate illustrations, sparking debates about the boundaries of creativity and copyright protection.^[5] While Kashtanova initially obtained a copyright registration for the entire novel, a subsequent review led to a reevaluation, ultimately resulting in a limited registration that excluded protection for individual AI-generated images. The Copyright Office clarified that there are ways for artists to incorporate AI-generated content into copyrightable works by introducing sufficient creativity. The Office acknowledged the copyrightability of Kashtanova’s original text and the “selection and arrangement of the text and images” together, but not the AI images independently. Additionally, it stated that “works containing otherwise unprotectable material that has been edited, modified, or revised by a human author can be registered, but in this case, there was insufficient evidence of such changes.”^[6] The USCO’s stance, rooted in the principle of human authorship, highlighted the perceived lack of substantial creative control exerted by Kashtanova over Midjourney’s output, underscoring the challenges of attributing authorship in collaborative AI artworks.^[7]

Following the partial cancellation of the registration for “Zarya of the Dawn,” Kashtanova pursued a new application for “Rose Enigma,” another artwork created with AI tool Stable Diffusion.^[8] Kashtanova’s deliberate efforts to guide the AI through specific prompts and creative decisions aimed to establish the artist’s substantial creative input and eligibility for copyright protection. This artwork, which is currently pending copyright registration exemplifies the ongoing struggle to reconcile AI-generated art with traditional copyright frameworks, as artists navigate the complexities of authorship and creative control in collaborative AI artworks.^[9]

In a separate instance, the USCO’s ruling on “Théâtre d’Opéra Spatial,” an AI-generated artwork by artist Jason Allen, further underscored the challenges of attributing human authorship to AI-generated works.^[10] Despite Allen’s involvement in providing prompts and shaping the final image, the USCO deemed the artwork ineligible for copyright protection, citing its non-human authorship. Allen clarified that he made numerous revisions and provided text prompts around 624 times to create the initial version of the image. Following Midjourney’s creation of the initial version, Allen then used Adobe Photoshop to eliminate imperfections and introduce new visual elements.^[11] Additionally, he utilized Gigapixel AI to enhance the image’s resolution and size. Despite the examiner’s request to exclude Midjourney’s contributions from the copyright claim, Allen chose not to do so. Consequently, the USCO declined to register the claim because the submitted work contained contributions from both Allen and Midjourney that were deemed inseparable. Allen subsequently appealed the denial and requested reconsideration from the USCO.

These decisions highlight the evolving legal landscape surrounding AI-generated art and raise critical questions about the scope and applicability of copyright law in the digital age. As artists continue to push the boundaries of creativity with AI, the need for clarity and adaptability in copyright frameworks becomes increasingly evident. Balancing the interests of human creators, AI developers, and the public requires thoughtful consideration of the unique challenges posed by AI-generated artworks.

The Challenges that Arise

The collaboration between human artists and AI presents a unique set of challenges for determining copyright protection, as it blurs the lines between human and machine creativity. One fundamental question that arises is the extent of human involvement required for a work to receive copyright protection. Since protection is only granted to works with a human author, determining the human-authored elements of AI-generated works poses significant difficulties, as AI algorithms can autonomously generate content based on predefined instructions and datasets.^[12] The distinction between human and AI-generated contributions becomes blurred, especially when the artist’s input involves selecting prompts, tweaking parameters, or guiding the overall creative process.^[13] This ambiguity makes it challenging to ascertain the level of creative control exerted by the human artist and the extent to which the final artwork reflects their original expression.

Moreover, even if a human artist modifies or enhances an AI-generated image, it may still be challenging to delineate the human-authored elements for copyright protection.^[14] The threshold for human intervention required to qualify for copyright protection remains unclear, raising questions about the sufficiency of minor modifications or alterations to AI-generated content.^[15] The USCO’s requirement for human authorship further complicates matters, as artists must demonstrate their substantial creative contribution to the final work to qualify for copyright protection and not something merely created by a “mechanical process” or simply inputting text prompts to generate art.^[16]

Furthermore, the question arises of whether a human work inspired by AI-generated content could potentially infringe upon the original source material.^[17] AI algorithms can be trained on vast datasets, including copyrighted works, raising concerns about the potential for unintentional infringement when creating derivative works. If a human artist’s work is based on AI-generated content trained on copyrighted material, it may inadvertently incorporate elements protected by copyright, leading to legal disputes and challenges to the work’s originality.

Addressing these challenges requires a nuanced understanding of the interplay between human creativity, AI technology, and copyright law. Clear guidelines and standards for determining the human-authored elements of collaborative AI art are essential to provide clarity and certainty for artists and creators. Additionally, robust mechanisms for identifying and addressing potential copyright infringement in AI-generated content are necessary to safeguard the rights of original creators and promote a fair and equitable creative ecosystem.

Art of Collaboration: The Skill of Working with AI in Art

It also may be important to note that while AI offers unprecedented possibilities for creativity, harnessing its potential may also require a unique set of skills and expertise from human artists. To create meaningful and impactful works with AI, artists must also navigate a complex interplay of technical proficiency, creative vision, and ethical considerations. A reddit user shared that they spent 30 hours on Midjourney learning how to use AI to create art and develop their skills.^[18] While some may not agree and argue that anyone can create AI art, it’s interesting to think of the skills that may be required to create more detailed or perhaps even meaningful artwork with AI.

At the heart of collaborating with AI lies the skill of effectively utilizing text prompts and other input mechanisms to guide the AI’s creative process. Unlike traditional artistic mediums where artists directly manipulate physical materials, working with AI involves crafting nuanced instructions and prompts to influence the output generated by the algorithm. This requires a deep understanding of the capabilities and limitations of AI tools, as well as a keen intuition for shaping the creative direction of the artwork. Furthermore, mastering the art of collaboration with AI entails proficiency in using AI platforms and tools to their fullest potential. From selecting the right algorithms and models to fine-tuning parameters and settings, artists must possess some technical acumen to navigate the intricacies of AI technology. Additionally, skills in editing and refining AI-generated content are essential for sculpting the final artwork into a cohesive and compelling piece of art.

Some may once again argue otherwise, since works can be generated even with a simple text prompt: “Midjourney allows anyone to generate unique, creative pieces with just a command. The only element that the AI art generator requires from a human being is a prompt: The user types in whatever they want, and the AI gives them four new images within 60 seconds.As long as you can access Discord, you’re good to go. Midjourney is in beta mode for public use.”^[19] However, creating different types of artworks may demand a high degree of creativity and artistic sensibility from human artists, especially if the AI art will be built upon by human artists in the process through hand-drawn images, like in the case of Rose Enigma or more detailed text prompts.

While AI algorithms can generate content autonomously, it is the human artist’s unique vision, aesthetic sensibility, and emotional depth that perhaps imbue the artwork with more meaning and resonance, which might be a factor to consider when thinking of copyright registration. Whether it’s crafting evocative text prompts, guiding the AI’s interpretation of visual concepts, or infusing personal narrative and symbolism into the artwork, human creativity remains indispensable in the collaborative process. Maybe anyone can generate an AI created artwork, but making it into something creative with AI could be a whole other story.

As an example to consider, comparing the use of AI to traditional creative tools like Photoshop highlights the distinct skill set required for each. While both platforms enable artists to manipulate and transform visual content, AI introduces a new dimension of complexity and nuance. Unlike Photoshop, which relies on manual input and manipulation, AI algorithms operate autonomously, requiring artists to adapt their creative process to interact with machine intelligence.^[20] This shift necessitates a learning curve and ongoing skill development to effectively use AI in artistic expression. Moreover, now Photshop softwares like Adobe also have AI software built in to assist with Photoshop as well, which shows the wide usage of AI for different purposes to edit and generate images. In essence, mastering the art of collaboration with AI could require a multifaceted skill set encompassing some level of technical proficiency and creative vision.

Ownership of Collaborative Works

When it comes to protecting their works in the context of AI collaboration, artists must carefully consider issues of ownership, authorship, and intellectual property rights.^[21] Firstly, it’s important to distinguish between authorship and ownership in copyright law. Authorship refers to the individual or entity that creates the work, while ownership refers to the legal rights and interests associated with that work.^[22] Generally authors of the work are also the owners of the work unless they sign away the rights to another entity. In the case of collaborative works involving AI, the question of ownership becomes particularly complex. AI platforms often assert ownership over the outputs generated by their algorithms, but the rights to these outputs may be transferred to the user who inputs the text prompts or provides other input. Analyzing the terms of various AI platforms sheds light on the implications for ownership and intellectual property rights:

OpenAI: OpenAI’s terms state that users retain ownership rights in their input and own the output generated by the AI. This grants users a degree of control over the content they create using the platform.^[23]

Midjourney: Midjourney’s terms stipulate that users own all assets created with the services, subject to certain exceptions. However, users grant Midjourney a broad license to reproduce, prepare derivative works of, and distribute the text and image prompts they input into the services, as well as any assets produced through the service. This raises concerns about the extent of control users have over their creations and the rights they retain.^[24]

Stable Diffusion: Users retain all rights, title, and interest, including any intellectual property rights, of the generated images. This agreement explicitly states that users do not grant ownership rights to the services, indicating a more favorable approach to user ownership.^[25]

These terms have significant implications for the ownership and protection of works created through AI collaboration. While some platforms may grant users ownership rights over their creations, others retain broad licensing rights that could impact users’ control and autonomy over their works. Issues may arise if users are not fully aware of the rights they are granting to AI platforms or if the terms are unclear or overly restrictive. Furthermore, determining ownership and authorship in collaborative AI works can be challenging, especially when multiple parties contribute to the creation process. Artists must carefully review the terms of AI platforms and consider the implications for their rights and interests before engaging in collaborative projects.

Protection of AI-collaborated works

Given the complexities surrounding the ownership and protection of collaborative works with AI, artists must adopt proactive strategies to safeguard their creations. Here are some steps artists can take to protect their collaborative works:

Use Watermarks and Digital Signatures: Artists can add watermarks or digital signatures to their works to establish ownership and authenticity. Watermarks can deter unauthorized use or reproduction of the work, while digital signatures provide a cryptographic seal of authenticity that verifies the creator’s identity.^[26]

Document the Creation Process: Keeping detailed records of the creation process, including text prompts, input parameters, and revisions, can serve as evidence of the artist’s creative contribution and ownership of the work. Documentation can help establish a clear chain of ownership and provide valuable insights into how and when the work was created. Perhaps as more registrations are looked over and decided by the copyright office or more lawsuits are heard and decided, documentation would be helpful in attaining any legal protection at a later stage.

Blockchain Registration: While blockchain registration does not replace copyright registration, it can provide additional evidence of ownership and authenticity. By timestamping and storing transaction data on a decentralized blockchain ledger, artists can establish a permanent record of their creations, including details such as the creator’s identity, creation date, and platform used. This can help mitigate disputes over ownership and provide proof of authorship in case of infringement claims.^[27]

Explore Alternative Licensing Models: Artists can consider alternative licensing models, such as Creative Commons licenses, to govern the use and distribution of their collaborative works. Creative Commons licenses allow artists to retain certain rights while granting others permission to use, share, or modify the work under specified conditions. Creative Commons’s General Counsel Kat Walsh has stated that “If you create works using generative AI, you can still apply CC licenses to the work you create with the use of those tools and share your work in the ways that you wish. The CC license you choose will apply to the creative work that you contribute to the final product, even if the portion produced by the generative AI system itself may be uncopyrightable.”^[28] By choosing a licensing model that aligns with their preferences and goals, artists can maintain control over their creations while promoting collaboration and sharing within the creative community.

By adopting these measures, artists can protect their collaborative works with AI and assert their rights as creators in the digital age.

Examples of AI-collaborative Artwork

Due to AI’s rapid development and advances in tech, there has been a notable surge in innovative collaborations between human creators and AI, with artists navigating legal frameworks or actively testing boundaries to explore AI’s potential as a creative partner. This dynamic experimentation underscores the excitement surrounding AI’s ability to inspire and elevate the creative process, offering fresh perspectives and innovative techniques to produce exceptional work. When approached ethically and responsibly, AI really could serve as an enjoyable collaborator, sparking new ideas and pushing artistic boundaries in ways previously unimaginable. Here’s a brief overview of how a few artists are collaborating with AI to create extraordinary pieces:^[29]

• Sougwen Chung: Explores human-machine relationships through interactive installations, incorporating AI in her creative process to delve into the intersection of technology, robotics, and the arts.
• Refik Anadol: Specializes in immersive installations, transforming large datasets into mesmerizing visual and auditory experiences using advanced algorithms for data visualization.
• Alexander Reben: Known for his “Blind Self-Portrait” series, Reben explores human behaviors through AI-generated artworks, including the world’s first TED talk written entirely by AI and presented as a cyborg.
• Sofia Crespo: Utilizes AI algorithms to create hybrid organisms and speculative life forms, challenging the boundaries between organic and artificial life through her digital artworks.
• Stephanie Dinkins: Explores issues of race, gender, and identity through conversational AI systems and virtual characters, engaging in dialogues around human experiences and consciousness.
• Ross Goodwin: Combines storytelling with computational creativity, blurring the lines between human and machine authorship, as demonstrated in his AI-generated screenplay and short film “Sunspring.”
• Jake Elwes: Explores identity and perception through surreal visuals generated by AI models trained on vast datasets, including his project “The Uncanny Archive” which generates portraits of influential figures in AI and technology.
• Taryn Southern: Collaborates with AI systems to create music compositions and performances, co-writing and co-producing her album “I AM AI” with AI algorithms, showcasing the fusion of human creativity and machine intelligence in the music industry.

Interview with Colleen Hoffenbacker on AI-collaborative Art and Copyright Registration

Through her vibrant floral oil paintings, American artist Colleen Hoffenbacker blends both traditional and digital mediums to explore the captivating interplay among humanity, nature, and technological progress. Following a successful career in commercial art and vintage poster restoration, she studied traditional realism at the Academy of Classical Design. Her work is recognized in the permanent Lunar Codex collections, the first curated art mission to include over 35,000 artists from around the globe on the surface of the moon. An award-winning member of American Women Artists, her artworks grace galleries, museums, and private collections worldwide, while also supporting the conservation efforts of environmental organizations through her artistic advocacy.

This interview was conducted by Atreya Mathur, Center for Art Law to understand Colleen Hoffenbacker’s artwork, the mediums she uses to create and her process of working with AI to create art, particularly her Regen and Floragen works as part of her ongoing RenAIssance and Ufloria series, which she received copyright registration for. Thank you to Hoffenbacker for taking the time and speaking with the Center about her work.

What kind of art do you enjoy creating? What are the mediums you usually use to create?

I specialize in oil painting, particularly exploring florals that gracefully balance on the edge of imaginative realism. Grounded in classical training, my approach emphasizes the rich, tactile qualities inherent in oil paint and is founded on an admiration for the interplay of light with natural forms. I also use digital 2D and 3D skills from my commercial art background to design custom references. Recently, I’ve integrated artificial intelligence into my digital process to harmonize the cutting-edge technology with timeless human values, creating a blend that’s uniquely mindful and full of wonder.

Where do you draw inspiration from to create?

I draw inspiration from visions of a synergistic coevolution between nature and technology, creating biophilic metaphors to align their delicate equilibrium, and using the visual power of AI to unveil the hidden, intricate, ethereal, and surreal beauty inherent in our natural world. Reflecting on the Renaissance era, I find parallels in their excitement for new artistic designs and mediums, such as oils, mirroring today’s enthusiasm for AI. The Renaissance’s reverence for nature as a teacher and belief in combining eclectic elements resonate with me, fostering a dynamic interplay between tradition and innovation in my work.

How did you come up with the series of artworks you created: (for the Regen and Floragen works)

This floral series is rooted in the conceptual fusion of enduring, nature-centric ideals from the past with the optimistic possibilities of today’s flourishing technological renaissance, using AI to reimagine and highlight nature’s beauty and diverse essence beyond traditional human observation. The metaphorical concept of generating with seeds inspires prompts that evoke botanical evolution, rebirth, and growth to guide the creations, providing fresh perspectives on nature. The integration of AI becomes not just a tool but a collaborator, contributing to the unfolding narrative of the artwork.

What was the process of creating these works? Can you let us know more about the AI software used and more about the original images first produced?

My preferred model is Stable Diffusion 1.5 and I use Photoshop, along with some upscaling software, for the manual editing and compositing of the initial input images to guide the model towards my vision as well as to create the final painting references. The Regen and Floragen works are part of my ongoing RenAIssance and Ufloria series, which are based on a single modified AI-generated source image with a Romanesque aesthetic, setting the tone for the output.

The generation process is quite fascinating—I generate AI-assisted preliminary reference images, not aiming for a single image of perfection but creating a photorealistic cache to work from. Many of them generally contain artifacts and are lacking in detail and composition considerations. The whole iterative process spans a few days, refining the floral prompt motif for consistency to generate the hundreds of raw material images and elements I will need to create a visually cohesive series of small paintings.

The two images above are the AI generated outputs which I combined in Photoshop to use as my Floragen 2.0.1 painting reference. The image below is a closeup detail of the final painting to show the composited flower transformed with oil paint onto an aluminum panel.

The resulting final amalgamated images become like my own version of stock photography, serving as inspirational springboards for my paintings. This is where my 20+ years of atelier training and experience come into play. I take these images and transform them through paint, gradually layering in nuanced details and bringing the scenes to vivid realism. It’s a fascinating blend of emerging technology and my classical artistic roots, and I’m excited to see how it continues to evolve.

These images above are closeups of the AI generated images compared with the final detailed results in oil paint for the Floragen 2.0.2 painting.

What was the process of copyright registration? Did you file for registration yourself or did you get in touch with an attorney?

I closely followed Kris Kashtanova’s and others’ AI copyright journeys. Though I did consider seeking legal advice, my background as a former graphic designer gave me a basic grasp of copyright laws related to images, so I decided to navigate the process independently. I was curious to see the outcome of forging my own path and opted to file for the copyright registration of these works on my own. It was a thrilling experience to receive the letter confirming their approval.

What would you say inspired this series? Did you have any goal or objective of sorts in mind for this artwork?

My studio is set in a beautiful high-tech region surrounded by forests, tulip farms, water, and mountains. I have always used digital technology as part of my creative process while feeling deeply connected with nature through decades of studio and outdoor painting. The Renaissance laid the groundwork for technological advancements by fostering a spirit of inquiry, exploration, and innovation. My goal is to create thought-provoking works that connect with viewers today and perhaps challenge traditional perceptions as we venture into a future realm shaped by AI.

Do you have any plans for what you may want to do with the art you have created?

As a female artist, I’m motivated by the increasing support for women in the art and tech spheres. My overarching goal is to contribute to the dynamic art scene and actively participate in essential conversations offering varied perspectives and cultivating a mindful stewardship of AI. A significant aspiration is to see my works housed in museums and tech art collections.

Conclusion

As artists continue to explore the delicate relationship between human creativity and artificial intelligence, the path forward is both exciting and nuanced. Through innovative collaborations and bold experimentation, creators are harnessing AI’s potential to push the boundaries of artistic expression. The artwork and collaborative process of creating the Regen and Floragen works by Colleen Hoffenbacker is a good example of combining AI-generated work with traditional oil paint to create collaborative AI-generated art which was also granted copyright protection. The skill of using AI along with traditional mediums of art to create artwork is an important factor to consider in the process of copyright registration with the USCO.

However, the journey for all artists is not without its challenges, regarding copyright issues, ownership concerns, and the delineation of human versus AI-generated contributions. Moving forward, it’s crucial for artists, AI developers, and legal experts to engage in dialogue and collaboration to address these complexities. Clearer guidelines and legal frameworks are needed to navigate the evolving landscape of AI-assisted creativity, ensuring that both human and machine contributions are recognized and protected appropriately. Additionally, increased transparency and accountability in AI platforms’ terms of service can empower artists to assert their rights and negotiate fair terms for collaboration.

Moreover, the integration of blockchain technology offers promising avenues for securing and verifying ownership of collaborative works, providing a decentralized and immutable record of creative contributions. By leveraging digital signatures, smart contracts, and blockchain-based registries, artists can potentially establish clear proof of authorship and ownership, enhancing trust and accountability in the creative process. As the world forges ahead into this new era of artistic collaboration, embracing the potential of AI while safeguarding the rights and integrity of creators will be paramount.

About the Author

Atreya Mathur is the Director of Legal Research at the Center for Art Law.

Select Sources:

1. What is Copyright, U.S Copyright Office, available at https://www.copyright.gov/what-is-copyright/ ↑
2. Id. ↑
3. Atreya Mathur, Art-istic or Art-ificial? Ownership and copyright concerns in AI-generated artwork, Center for Art Law (Nov. 2022), available at https://itsartlaw.org/2022/11/21/artistic-or-artificial-ai/ ↑
4. United States Copyright Office, Registration #VAu001480196, available at https://www.copyright.gov/docs/zarya-of-the-dawn.pdf ↑
5. Id. ↑
6. Stephen Wolfson, Zarya of the Dawn: US Copyright Office Affirms Limits on Copyright of AI Outputs, Creative Commons (February 27, 2023) available at https://creativecommons.org/2023/02/27/zarya-of-the-dawn-us-copyright-office-affirms-limits-on-copyright-of-ai-outputs/ ↑
7. See Atreya Mathur, Copyright Registration for AI Generated Works: Zarya of the Dawn and the Dawn of a New Creator, Center for Art Law (Jul. 6, 2023) available at https://itsartlaw.org/2023/07/06/copyright-registration-for-ai-generated-works-zarya-of-the-dawn-and-the-dawn-of-a-new-creator/ ↑
8. Registration Cover Letter for Rose Enigma, available at https://drive.google.com/drive/u/0/folders/1MOZe5wLRJF3hR6I-x77GXeuTG-QAUv37 ↑
9. See Atreya Mathur, A Rose Enigma: Pending (Right) Protection, Center for Art Law (Dec. 8, 2023) available at https://itsartlaw.org/2023/12/08/rose-enigma/ ↑
10. United States Copyright Office, Second Request for Reconsideration for Refusal to Register Théâtre D’opéra Spatial (SR # 1-11743923581; Correspondence ID: 1-5T5320R), available at https://www.copyright.gov/rulings-filings/review-board/docs/Theatre-Dopera-Spatial.pdf ↑
11. Amol Parikh, Tragic Ending: Award-Winning AI Artwork Refused Copyright Registration, IPupdate (Sept. 21, 2023) available at https://www.ipupdate.com/2023/09/tragic-ending-award-winning-ai-artwork-refused-copyright-registration/ ↑
12. Pavan Vadapalli, Generative AI vs Traditional AI: Understanding the Differences and Advantages, upGrad (Aug. 15, 2023) available at https://www.upgrad.com/blog/generative-ai-vs-traditional-ai/ ↑
13. Arthur Roberts, et.al., Who owns the copyright in AI-generated art?, Murgitroyd Blog (2022), available at https://www.murgitroyd.com/en-us/blog/who-owns-the-copyright-in-ai-generated-art/ ↑
14. Id. ↑
15. See Atreya Mathur, Art-istic or Art-ificial? Ownership and copyright concerns in AI-generated artwork, Center for Art Law (Nov. 2022), available at https://itsartlaw.org/2022/11/21/artistic-or-artificial-ai/ ↑
16. Id. ↑
17. Alex Springer, Publishers’group warns that generative AI content could violate copyright law, Marketing Brew, available at https://www.marketingbrew.com/stories/2023/06/05/publishers-group-warns-that-generative-ai-content-could-violate-copyright-law ↑
18. See AI art requires a lot of skill on my part, Reddit, available at https://www.reddit.com/r/lies/comments/1765pgl/ai_art_requires_a_lot_of_skill_on_my_part/ ↑
19. Sherin Shibu, I tested out an AI art generator and here’s what I learned, zdnet (Sep. 27, 2022), available at https://www.zdnet.com/article/i-tested-out-an-ai-art-generator-and-heres-what-i-learned/ ↑
20. See Katie Terrell Hanna, What is Photoshop, techtarget, available at https://www.techtarget.com/whatis/definition/Photoshop#:~:text=It%20uses%20a%20layer%2Dbased,as%20a%20tool%20for%20professionals. ↑
21. Atreya Mathur, Art-istic or Art-ificial? Ownership and copyright concerns in AI-generated artwork, Center for Art Law (Nov. 2022), available at https://itsartlaw.org/2022/11/21/artistic-or-artificial-ai/ ↑
22. The Compendium of the U.S. Copyright Office Practices (3d.), available at https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf ↑
23. See OpenAI Terms of Use, available at https://openai.com/policies/terms-of-use ↑
24. See Midjourney Terms of Service, available at https://docs.midjourney.com/docs/terms-of-service ↑
25. See Stable Diffusion Terms and Conditions, available at https://stablediffusionapi.com/p/Terms-and-Conditions ↑
26. Atreya Mathur, Patrick K. Lin, Irina Tarsis, U.S. Copyright Office, Comment from Center for Art Law (Nov.1, 2023) available at https://www.regulations.gov/comment/COLC-2023-0006-8315 ↑
27. Id. ↑
28. Kat Walsh, UNDERSTANDING CC LICENSES AND GENERATIVE AI, Creative Commons (Aug. 2023) available at https://creativecommons.org/2023/08/18/understanding-cc-licenses-and-generative-ai/ ↑
29. See the list of artist: Karen Garces, 12 AI Artists Who Exemplify the Weird World of AI Art, penji (Jul. 29, 2023) available at https://penji.co/ai-artists/ ↑