You may remember the hubbub and hullabaloo that ensued when Josh Allen, a digital artist, won first place at the 2022 Colorado State Fair’s Fine Arts Competition in the Digital Arts/Digitally Manipulated Photography contest with an image—”Théâtre D’opéra Spatial”—that he had created using AI. Allen continues defending his work, now butting heads with the Copyright Office.
Let’s take a few moments to examine the situation and its potential ramifications in the future.
Copyright Office: “No.” Allen: “Not So Fast…”
In the fall of 2023, Allen took his prize-winning piece and attempted to secure copyright registration. The Copyright Office refused twice, stating that there simply wasn’t enough human authorship in prompting an application to generate an image to secure the protections Allen was seeking. As the Copyright Office Examiner’s conclusions stated, Allen had “no control over how the artificial intelligence tool analyzed, interpreted, or responded to these prompts.”
Allen Has an Assortment of Arguments Against this Ruling
Josh Allen is appealing the decision and requesting a judicial review, arguing that the Examiner was swayed by the public backlash and discourse his work had caused… particularly in the media.
As such, Allen has argued that the Examiner considered improper factors while ruling and questioned whether an Examiner could even tell whether art was AI- or human-generated. This would make any rules requiring an Examiner to make this differentiation arbitrary. By denying his copyright claim, Allen also argues that the Copyright Office is setting a dangerous precedent concerning AI-generated content and its ownership.
As AI and the technology that powers it improve, determining who has authorship rights will only become more challenging. Each time the Copyright Office makes a mistake, a dispute will likely take up valuable time in the courts.
Allen can now make his case to a jury, arguing that the Copyright Office failed to consider the amount of human authorship his process and final work ultimately contained.
Allen’s stance is that, rather than using Midjourney (an AI platform) to generate random outputs, he went about using it as a tool to generate a specific image he had stuck in his head—a tableau of “women in Victorian dresses wearing space helmets” as they were “performing opera on stage,” with “their attire presenting a juxtaposition between old-world charm and a futuristic twist.”
To accomplish this, Allen reportedly spent over 100 hours crafting over 624 prompts, all to determine which instructions worked best so that he could direct Midjourney into creating the image, so to speak. As he puts it, the time and effort (and frustration) he poured into this project surpasses the standard copyright law sets for human authorship, as he guided and directed the AI platform into fulfilling his vision.
To counter this argument, the Copyright Office has agreed that Allen’s prompts are indeed copyrightable, but not the results that Midjourney generated. Furthermore, the Copyright Office has stated that if he had altered the results a little more, such as by applying a filter, registering his copyright would have been a non-issue.
Time Will Tell How this Case Pans Out
Considering that the drama surrounding “Théâtre D’opéra Spatial” first went viral a few years ago, and the Copyright Office refused to register its copyright just last year, it seems that this saga is far from over.
Regardless of the outcome, however, these events have shown how quickly technology can become complicated, both in its implementation and the legality surrounding it. Jason Allen has spent years trying to establish his claim to artwork that, should he win his case, is his. In essence, he’s arguing that he is the creator of a certain data set, using the assistance of AI to compose it.
So, what do you think of all this? Is AI fair game to be used as an artist’s tool, protectable under copyright, or is the fact that a machine does such a significant portion of a task enough to make such a claim moot?