On September 5, 2023, the Review Board of the US Copyright Office issued a decision refusing to register “Theatre D’Opera Spatial,” a two-dimensional artwork by Jason M. Allen. The image had drawn national attention in 2022 as the first AI-generated work to win the Colorado State Fair’s annual fine art competition.
Allen created the image with Midjourney, a text-to-image AI service, then refined it in Adobe Photoshop and upscaled it with Gigapixel AI. According to the decision, he stated he had “input numerous revisions and text prompts at least 624 times to arrive at the initial version of the image.” The Office concluded the work contained “more than a de minimis amount of content generated by artificial intelligence,” which therefore had to be disclaimed in any registration. Because Allen refused to disclaim the Midjourney-generated portion, the Office held the work could not be registered as submitted.
The Board reasoned that Allen’s prompting did not make him the author of the Midjourney image, because the system - not the user - determined how the prompts became a picture. It quoted prior rulings that human authorship is “a bedrock requirement of copyright” and noted Midjourney “does not interpret prompts as specific instructions to create a particular expressive result.” The Board left open that Allen’s Photoshop edits might themselves be copyrightable, but he had to limit his claim to those human contributions.
Why business readers should care: the decision set an early, concrete line - typing prompts into an AI image generator does not, by itself, make you the author for copyright purposes, even when the output wins an art prize.