Copyright Protection for Works Generated by AI

Within the past few years, artificial intelligence, or generative AI, a new technology created to enable computers to simulate human capabilities, has gained popularity as a tool used by many individuals and companies within their professional work (Stryker, 2024). Specifically, some artists and writers have begun to use AI within their artistic products. The inclusion of AI now brings into question the eligibility for copyright protection for these creations. Copyright protection is available for those who produce original works, including written and artistic creations. With the addition of AI, creative works involving AI generation should be protected via copyright to a clearly defined, limited extent.

Ineligibility for Copyright Protection

The situations in which the use of generative AI should not result in copyright protection include when another work is used as inspiration directly in the prompt given to the AI and the manner in which the prompt is written. In the United States, copyright protection is available “in original works of authorship fixed in any tangible medium of expression, now known or later developed, from which they can be perceived, reproduced, or otherwise communicated” (U.S. Copyright Office, 2024). Thereby, works that are not considered original works are not eligible for copyright protection. Original is defined by the U.S. Copyright Office as “independently created by a human author and ha[s] a minimal degree of creativity” (U.S. Copyright Office, 2025). By these standards, works that incorporate generative AI would not be considered original works, since generative AI uses data from online sources and existing material to generate results, lacking the degree of creativity. These facts lead to necessary restrictions on generative AI copyright protection. Works that should not be eligible for copyright protection are also based on the prompts presented to the generative AI. When a human inputs a prompt, the wording, emphasis, and other factors of the prompt affect the results from the AI. It would not be original, then, for a human author to give AI an image, which is under copyright protection with the rights owned by another human, and ask it to ‘create an image similar, but change the insert aspect,’ since the resulting image would be a near copy of the already protected original work, thereby making it not an original work. Therefore, the prompt given would be the catalyst for defining the originality of the work.

AI As a Tool

One can argue that the definition of original also states that original work must be independently created by a human author; therefore, with the assistance of generative AI, works are not independently created and are also not created by a human author, since AI is generating the work. However, these arguments are incomplete when one looks deeper into these meanings. Since the prompt is the catalyst for the AI to generate, and the phrasing and diction have a direct impact on the results, the human author is the direct source from which the generation is stemming from, leading to unique results for that human user. Furthermore, despite the use of generative AI being seen as an aid, therefore not being independently created, AI is rather a tool similar to other digital programs, such as the Adobe Platforms or even Canva, which help artist create their own unique artistic visions and original pieces. AI can be categorized as a tool because it serves to enhance human capabilities rather than serving as an independent entity. Since generative AI would not be considered an independent entity, but rather a tool, artists or writers who utilize generative AI for enhancing their work and capabilities should be eligible for copyright protection.

Eligibility for Copyright Protection

Although there may be some reservations about allowing copyright protections for works with generative AI, some situations stand out the most, where copyright protection would be necessary for the creator or author. The situations in which works are eligible for copyright protection include when the AI-generated work is edited or changed to a certain extent by the original prompt writer, when AI is used on an existing original piece from the prompt writer, and when AI is only taking inspiration and information from works in the public domain. Firstly, copyright protection should be established for work where AI was used first and altered by the original prompt writer because the whole work was not solely created by the generative AI, but rather has the addition of originality when the human who prompted the AI creation alters the given result. The piece, when changed by the human creator or author, directly becomes an original work of that person, even if it was changed with minimal creativity. As mentioned previously, copyright protection should be established for work where AI is used on an existing original piece from the prompt writer, since the generative AI would be a tool to enhance the work rather than as a secondary aid in creating the work. Finally, copyright protection should be established for work where generative AI’s sourcing comes from works in the public domain and not those with copyright protection. Since copyright protection applies to various works and one cannot legally breach copyright, then generative AI should also be under those restrictions when being used as a tool for writers, artists, filmmakers, or songwriters. Generative AI uses information from data sets to be trained to produce a quality output. For example, to generate images, AI uses a data set of photos to train to produce a quality image for its users. Therefore, generative AI should be restricted to a data set that contains images, songs, videos, and more that are within the public domain. Having this standard in place would ensure that the humans who use generative AI in their work would be able to avoid infringing on works with copyright protection.

Who Should Own the Rights?

With the situations eligible for copyright protection clearly defined, one must now answer the question: Who should own the rights when generative AI has been used? The answer to this question is simple: the human who wrote the prompt given to the AI, known as the prompt writer, should own the rights, given that the work prompted by the AI contributed to the prompt writer’s already existing work, or the prompt was given first and then altered by the prompt writer themselves. As established earlier, the prompt given defines the results that generative AI creates. Since the human author or artist writes the prompt, which leads to the results given by AI, then the human who set or made the prompt should own the rights to the generated work.  

One can argue that the creators of the AI should own the rights or that the AI itself should own the rights; however, the creators of the AI have simply supplied the AI with the data sets to train the AI to be able to generate the information requested by the user. The creators play little to no role in the originality of the work that generative AI was used for; therefore, those creators should not own the rights. Further, the AI itself should not own the rights because firstly, AI is not human, and secondly, AI was created to not work as an independent entity; therefore, without a human to initiate the creation, the AI would not create the result of the prompt.

Conclusion

Generative artificial intelligence has vastly impacted the sphere of creative works, including writing, songwriting, editing, video creation, photography, and art. This new technological addition calls into question many new legal aspects, which include copyright protection and ownership of work. The use of generative AI in these types of works should have copyright protection within specific instances and with specific restrictions, and the owner of these rights should be the human prompt writer. As AI evolves, these restrictions and specifications may also evolve; however, at the level generative AI is at now, the copyright protections suggested should begin to be established.