General

Who owns the copyright for AI-generated Thanksgiving recipes? – Briefing on American University Intellectual Property

Who owns the copyright for AI-generated Thanksgiving recipes?  - Briefing on American University Intellectual Property

New York Times food writer Priya Krishna used OpenAI products to generate new Thanksgiving recipes and images, which begged the question: who owns the copyright for these recipes? According to OpenAI’s terms of use, Ms. Krishna is the owner, but in reality, copyright for machine-generated content is more complicated than that.

Enforcing copyright protection for recipes is notoriously difficult. While US copyright law strives to “original works of authorshipBy excluding unauthorized reproduction of works of authorship, including literary works, pictorial and graphic works, dramatic works, etc., this protection does not extend to ideas, procedures or processes described or explained in the work. 17 U.S.C. § 102. When a recipe is copyrighted, the copyright only applies to the literal text of the recipe, so the recipe can be reproduced with minor changes without infringing copyright protection. To prevent recipe theft, recipe developers and authors have developed permissions with instructions for reusing their work. Cookbook authors have started adding more easily protected elements to cookbooks, including narrative essays and photography, both of which can be copyrighted and less easily reproduced without infringement. While copyright protection of recipes can be somewhat ineffective, AI-generated recipes still pose an interesting question: who owns these recipes? The AI? The AI ​​developers? Or the user entering the recipe prompts?

In anticipation of Thanksgiving, New York Times food writer Priya Krishna experimented with artificial intelligence to create new Thanksgiving recipes. She used OpenAI products GPT-3 and DALL-E to create and write the new Thanksgiving recipes and generate images of the recipes. To write the recipes, she used OpenAI’s language prediction model GPT-3, which stands for Generative Pre-trained Transformer 3. GTP-3 is a neural network designed to receive and understand language input and transform the input into what is predicted to be the most useful language output for the user. It has gotten about 570 GB of text information from multiple publicly available sources, including Wikipedia, and is the largest neural network ever created. DALL-E works similarly to GPT-3, with an input and output model. For both, the “inputs” are the language, commands, reference materials that are fed into the service (GPT-3 or DALL-E), the service then generates and returns output based on the input received. In the video, Mrs. Krishna shares her culinary background and preference for recipe development. GTP-3 used these inputs to create recipes for new Thanksgiving dishes, including pumpkin spice chaat and naan stuffing. Krishna then used the GPT-3 generated recipes and inputs to the image generator DALL-E, which generated eerily realistic renderings of the recipes.

OpenAI terms of uselast updated November 3, 2022, states that users own their “inputs” for both DALL-E and GTP-3. In the Terms of Use, Open AI has assigned the right, title, and interest of OpenAI in and to the User Generated Output to the User.. So, according to OpenAI, Priya Krishna “owns” the recipes and photos generated by GPT-3 and DALL-E through her input. However, Open AI is quick to note that GPT-3 and DALL-E, while generating new and original content, can still generate the same or similar output for OpenAI users or third parties.

OpenAI’s terms of use state that Priya Krishna owns the recipes GPT-3 and photos DALL-E generated under her input; however, USCO has denied copyright of AI-generated works. On February 14, 2022, USCO denied copyright of an artistic work titled “A Recent Entrance to Paradise” generated by AI. In their reasoning, the board ruled that autonomously generated creative works without creative contributions from a human actor do not meet the standard for copyright protection. USCO stated that copyright law only protects “‘the fruits of intellectual labor’ which are ‘based on the creative powers of the [human] mind.’ COMPENDIUM (THIRD) § 306 (citation Trademark Business, 100 U.S. 82, 94 (1879)).” In this case, the applicant’s intention was to fully credit AI for the creation of “A Recent Entrance to Paradise”, so the applicant did not include their human input as a basis for copyright in the registration application. USCO declined to comment which level of human involvement in the creation of machine-generated works would meet legal criteria for copyright protection. The Supreme Court and lower courts have rejected attempts to extend copyright protection to non-human creations. Federal agencies have followed the courts and ruled that while devices can be used in creating creative works, there are “at least minimal human creative effort at the time the work is produced.If the applicant in the “A Recent Entrance to Paradise” case included his contributions to the creative work instead, USCO might consider the creative work to be the result of human creativity. However, the courts and USCO have yet to say how much “human creativity” it would take to grant an AI creation copyright protection.

In the video, Krishna used very generic commands, so it is unlikely that USCO would discover that her input generated output eligible for copyright protection. But what would happen if she uploaded her entire cookbook as input? Would that meet the legal requirements for human involvement in machine-generated work? USCO has yet to face this issue. With the growing prevalence of AI across industries, it is only a matter of time before USCO is forced to address more of these pressing questions for the protection of AI-generated creative works.