On February 21, 2023, the Copyright Office eclipsed its previous AI authorship decisions when it partially canceled Kristina Kashtanova’s registration for a comic book titled Zarya of the Dawn. In doing so, the Office found that the AI program used by Kashtanova – Midjourney – was primarily responsible for the visual output that excluded the Office from Kashtanova’s registration. (Midjourney is an AI program that creates images from text descriptions, similar to OpenAI’s DALL-E.) The decision underscores not only the tension between copyright law’s requirements for human authorship and the means of expression authors can use, but them also raises the question: can AI-generated works ever be protected by US copyright law?
The decision
On September 15, 2022, the US Copyright Office issued a registration for Kashtanova’s comic book Zarya of the Dawn, which consists of text written by Kashtanova and images created with Midjourney. The office then learned of Kashtanova’s public statements that she used an AI program to create Zarya of the Dawn, caused the registration to be deleted, and gave Kashtanova an opportunity to explain why the registration should not be deleted. In response, Kashtanova argued that she used Midjourney the way a photographer uses a camera or a graphic designer uses Adobe Photoshop, and that her “core creative input” – i.e. text “prompts” and “pre-developed” images – is being used as such, as is her “Iterative process” of selecting midjourney images led to her creation of Zarya of the Dawn.
While the Office determined that Kashtanova’s selection and arrangement of text and images in the Zarya of the Dawn comic was protectable, it determined that Kashtanova did not author the actual imagery in the comic because Midjourney generates images in an “unpredictable manner”, such as z that Kashtanova did not “really sculpt” these images. The Office based its conclusion on “the significant gap between what a user can direct Midjourney to create and the visual material Midjourney actually produces”, noting that “Midjourney users do not have sufficient control over the images produced have to be treated as the ‘mastermind’ behind them. The bureau also dismissed Kashtanova’s argument that she authored the images through her “creative, human-written prompts,” since prompts in Midjourney are merely “suggestions,” not “commands.”
analysis
While this decision is consistent with the human authorship requirement reflected in previous Office and judicial decisions, it raises several important questions.
First, if copyright authorship is nullified by the mere use of “unpredictable” tools, how does this affect experimentation in art? In order to create works of art – many of which are widely regarded as the “core” of copyright protection – artists routinely take risks and experiment with different tools, mediums and creation methods that are by definition unpredictable. How many of the most important works of art would be refused protection if judged on the predictability of the means of expression used? If we follow the Office’s reasoning to its logical conclusion, would this lead to untenable results?
Second, although the Office came to the same conclusion as for all midjourney images, arguing that midjourney users lack creative control, can users exercise a greater degree of control over midjourney images? For example, Kashtanova generated some images with text prompts and others with “pre-developed” images. Did Kashtanova exercise more creative control when providing visual content for Midjourney to emulate? And could some users’ text prompts be detailed and nuanced enough to cross the line from “suggestion” to “order”?
Third, how much creative control is sufficient to support authorship under US copyright law, and to what extent does the creative control required vary with the means of creative expression used? While the Office is accustomed to deeming photographs eligible for protection and therefore unlikely to question whether a photo was taken with creative control over lighting, angles, timing, etc., its decision on Zarya of the Dawn suggests that AI -generated works are subjected to stricter scrutiny. Such a review may be unjustified; for example, one could argue that Kashtanova exercised the same degree of control over midjourney that a film director exercises on set, where there can often be improvisation by the actors and actresses. Assuming that Kashtanova made significant contributions to the imagery in Zarya of the Dawn, does she deserve co-authorship less than a film director? See Garcia v. Google, Inc., 766 F.3d 929, 933 (Cir. 9, 2014), on reh’g en banc, 786 F.3d 733 (Cir. 9, 2015) (discussing joint authorship in connection with a film. )
A New Dawn for Copyright in AI-Generated Works?
While the Office’s decision on Zarya of the Dawn may not bode well for copyright protection of AI-generated works, it may also be an outlier. Kashtanova’s statement of sole authorship was inconsistent with her public statements about Midjourney, and her use of “Zendaya” in text announcements suggests that Zarya was a famous copy rather than an original creation. In addition, the Office limited its decision to mid-journey, noting that “[i]It’s possible that other AI offerings that can generate expressive material work differently than Midjourney.”
So, obviously, the broader question of copyrightability of AI-generated content remains open. Creators, developers, and content platforms should pay careful attention as these issues continue to seep through the Bureau and the federal courts, and that copyright issues surrounding AI-generated works continue to evolve and hopefully be resolved.
[View source.]
