BLOG POST: Looking Through the Lens: Can Photography’s Past Support the Copyrightability of AI Art?

2024Copyright

Emily Luu

*This writing is a blog post. It is not a published IPTF Journal article.

The public release of text-to-image models, such as Midjourney and DALL-E, has allowed users to easily create AI-generated images using only a few words.[1] Text-to-image models take user-generated prompts and, within seconds,  produce an image matching that description.[2] Although text-to-image programs are a new type of visualization tool that significantly increase accessibility to artistic works for artists and non-artists alike, they have also drawn criticism from traditional artists who view this innovation only as a tool used for perpetuating uncreativity and “copycat” works.[3]As of March 2023, the U.S. Copyright Office has legally excluded AI-generated works from copyright protection because AI fails to meet the human authorship requirement of copyright law.[4] However, current artists who utilize AI-generated images in their work seek to challenge this rule as evolving technology redefines art.[5] In response, the Copyright Office is undertaking an examination of copyright law and policy issues raised by AI systems, including a consideration of comments from various stakeholders such as artists that both do and do not use AI generated images in their artwork.[6] This investigation has received over 10,000 comments, a number far greater than what is typically seen on other topics.[7]The Copyright Office plans to issue a report in 2024, which will contain several sections analyzing the various issues associated with AI-generated art.[8]

As the art community watches these new changes unfold, examining the development of photography offers a more positive outlook on the impact disruptive technology has on the art industry.[9]The context in which photography became recognized as an artform by the general public and protected by copyright law occurred organically and concurrently.[10] The general consensus that photography is in fact a form of art coincided with the legal recognition of photography in 1884 when the Supreme Court held in Burrow-Giles Lithographic Co. v. Sarony that photography is copyrightable. The Court ruled that a photograph itself does, in fact, incorporate the original intellectual conceptions of the photographer despite the lack of originality or novelty in operating the camera.[11] Furthermore, sufficient originality and creativity is displayed in the framing and posing of the subject.[12] The result of photography’s acceptance into the art world, both socially and legally, was the growth of art itself; artists were no longer limited to copying the real world because photography could capture reality for them.[13]

Today, AI-generated art is experiencing a similar entrance to the art industry: on one hand, the general public has embraced its novelty and the accessibility it gives to non-artists, while on the other hand some artists have pushed back on use of this new technology because they fear the text-to-image models are trained on stolen images resulting in theft of artist-created work.[14] However, some established artists have utilized text-to-image models to create new work of their own.[15] Although most photographs are now copyright protected, the Copyright Office has determined that works involving AI-generated images without substantial human authorship do not meet the threshold requirements for copyright protection because AI-generated art lacks human control in its image-creating mechanism.[16] But, some argue that, like a photographer who has control of the composition of the work and elements through framing, lighting, and choosing the subject, an artist using a text-to-image model has sufficient control of the composition through writing their own prompt and making edits to the generated image through image-editing programs like Photoshop.[17]

The Copyright Office should be open to expanding the definition of “human authorship” in light of this ever-advancing artificial intelligence technology.[18] Photography’s early enjoyment of copyright protection provided established and aspiring photographers great freedom to explore the art medium without fear that the fruits of their intellectual labor be stolen, leading to significant progress in the arts in alignment with the ultimate goal of copyright law.[19] Limitations should still be in place to address the criticisms of AI-generated art, but copyright law must adapt and expand to reflect the rapid technological progress being made with artificial intelligence.[20]


[1] Rachel Gordon, 3 Questions: How AI Image Generators Work, MIT CSAIL (Oct. 22, 2022), https://www.csail.mit.edu/news/3-questions-how-ai-image-generators-work (discussing the rise of AI image generators).

[2] See id. (discussing what AI image generators are and how text-to-image models work).

[3] See Beatrice Nolan, Artists Say AI Image Generators Are Copying Their Style to Make Thousands of New Images — And It’s Completely Out of Their Control, INSIDER (Oct. 17, 2022), https://www.businessinsider.com/ai-image-generators-artists-copying-style-thousands-images-2022-10 (discussing the widespread use and accessibility of AI image generators to the general public). As of 2022, DALL-E had more than 1.5 million users alone. Id.

[4] Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence, 88 Fed. Reg. 16,132 (Mar. 16, 2023) (to be codified at 37 C.F.R. pt. 202) (discussing the standards set by the US Copyright Office (USCO) in addressing the copyrightable factors of a work that includes AI generated images).

[5] See Adam Shrader, Another A.I.-Generated Artwork Was Denied Copyright Protection, Adding a New Knot to the Complexities of Creative Ownership, Artnet (Sept. 26, 2023), https://news.artnet.com/art-world/ai-art-copyright-2367590 (discussing the general controversies surrounding AI image generators, including “orphaned” art and inventions that will lack legal protection and be subject to stolen credit).

[6] See Craig Clough, Copyright Office’s GC Promises Report on Generative AI, LAW360 (Nov. 15, 2023), https://www.law360.com/articles/1767348/print?section=california (discussing the current updates on the USCO’s study on generative AI).

[7] Id.

[8] Nora Scheland, Looking Forward: The U.S. Copyright Office’s AI Initiative in 2024, Library of Cong. Blogs (Mar. 26, 2024), https://blogs.loc.gov/copyright/2024/03/looking-forward-the-u-s-copyright-offices-ai-initiative-in- 2024/.

[9]  See Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53, 58 (1884) (finding that photography was not included under original copyright statutes because the technology did not yet exist at the time of drafting or enactment).

[10] See Naomi Rosenblum, Andy Grundberg & Beaumont Newhall, History of Photography,Encyclopedia Britannica (Aug. 17, 2023), https://www.britannica.com/technology/photography (discussing the copyright decisions regarding photography in its historical context).

[11] Sarony, 111 U.S. at 61 (holding that a photograph is subject to copyright protection on the basis of originality and human authorship in the creation of the photograph).

[12] Id. at 60.

[13] See Rosenblum, Grundberg & Newhall, supra note 11 (discussing the rise of photographic societies that promoted photography as a serious art form through innovations in composition).

[14] Nolan, supra note 4.

[15] Id.

[16] See Clough, supra note 7 (discussing the role of randomness and lack of control in commercially available AI systems on the Copyright Office’s decision to limit copyrightability of AI-generated images based on the human authorship requirement).

[17] See Mannion v. Coors Brewing Co., 377 F. Supp. 2d 444, 450 (S.D.N.Y. 2005) (highlighting the role of human authorship in deciding the elements of originality in a photograph); Shrader, supra note 6 (describing how an artist using AI generative images in a piece controlled the composition of his work through specific selections of the image to adjust tone of the piece).

[18] See 17 U.S.C. § 102(a) (noting that Congress may broaden the reach of the statute to mediums “now known or later developed”).

[19] See Rosenblum, Grundberg & Newhall, supra note 11 (discussing new photographic art forms, including naturalistic photography, documentary photography, and photojournalism).

[20] See 17 U.S.C. § 102(a) (noting that the extent of copyright law is not limited to the scope of present technology and is open to expanding to recognize forms as creative and worthy of protection)