On Stephen Thaler’s Deeply Harmful Quest for AI Copyright
The Supreme Court is about to hear a case on copyright, but it’s not what you think.
<p dir="ltr">In the fall of 2023, <a href="https://www.theartnewspaper.com/2023/09/02/artificial-intelligence-lawsuit-decision-us-copyright-law">I wrote an article for The Art Newspaper</a> debunking the common misconception that AI-generated works of art were not eligible for copyright protection.</p>
<p dir="ltr">The United States Copyright Office is outdated, but the tenets that guide its decisions have largely proven to be adaptable, even though they haven’t really been updated since before the invention of the Internet.</p>
<p dir="ltr">When it comes to Artificial Intelligence and copyright, there are several crucial misconceptions. The greatest of them is the misquoted notion that AI-generated works are not eligible for copyright. This is a very common misunderstanding, fueled by click-bait articles, mostly due to one man’s rather arrogant crusade to get copyright protection for works generated autonomously by his AI.</p>
<p dir="ltr">Stephen Thaler’s various machines and algorithms have been the subject of much debate. The work that is at the center of the ongoing lawsuit, A Recent Entrance to Paradise, by Thaler’s Creativity Machine in 2012, was <a href="https://www.copyright.gov/rulings-filings/review-board/docs/a-recent-entrance-to-paradise.pdf">denied copyright several times</a>, and will now be considered by the United States Supreme Court.<strong> </strong></p>
<p dir="ltr">A Recent Entrance to Paradise, although conceptually interesting, is woefully… clunky, as far as AI-generated works of art are concerned, not to mention aesthetically unresolved. I regret, on an art historical level, that we cannot be debating this topic with an artwork that is at the very least more intriguing than a photograph of a French backyard. That it was made 13 years ago certainly makes one view it with kinder eyes, as does the fact that the first petition for copyright was made in 2018, before AI-generated works of art took over our feeds and our lives.</p>
<p><strong><img src="https://storage.googleapis.com/taex-storage-prod/news/upload-2070342605.jpg" alt=" A Recent Entrance to Paradise, a work produced by Dr. Thaler’s Creativity Machine" width="640" height="480"> <br><span class="custom-span">A Recent Entrance to Paradise, the work produced by Dr. Thaler's Creativity Machine</span></strong></p>
<p dir="ltr">Both in my conversations with them and in their rulings, the USCO insists that any AI generated material submitted for copyright consideration be made with a guiding human hand. But where I think a human hand is needed for aesthetic reasons, the USCO has much more practical concerns in mind: </p>
<p dir="ltr">The ultimate purpose of copyright is to know who should receive profit, if there is any to be had. Thaler is arguing that he grandfathered A Recent Entrance to Paradise, as the inventor of the Creativity Machine, and if granted copyright, he would be the beneficiary. A bit tenuous, but his argument is logical for the purposes of copyright debate. The issue here is that even if he were to succeed in this Don Quoixotean (as opposed to Sisyphean) quest, we do not yet live in a world where algorithms or machines have their own, independent bank accounts with which to participate in the free market economy. </p>
<p dir="ltr">Thaler’s insistence could set a precedent for machines to have their own streams of revenue through art and invention in the long run. But that is not the world we are living in today. I also think there is something to be said for a riveting court case, maybe 100 years from now, where a robot passionately argues for copyright eligibility for its own creations, rather than a man who is doing it for what I can only consider to be narcissistic reasons. After all, not a single AI-artist in my orbit is worried about obtaining copyright, and neither are most traditional artists.</p>
<p dir="ltr">Thaler’s insistence that autonomous AI-generated works of art merit copyright is too niche and exhausting a battle to be waged so publicly, when human-driven AI-generated works of art have yet to be fully accepted. But because there is only one high-profile court case being covered in mainstream news, most people are under the impression Thaler’s case is about the copyrightability of all AI-generated works. It isn’t.</p>
<p dir="ltr">The real damage here is the widespread misquoted news of Thaler’s rejections, which then lead people to think AI-generated works in general are not eligible for copyright. </p>
<p dir="ltr">This not only hurts the perception of the technology and its participants, it wounds the very nascent field of AI-generated art by framing all of us as illegitimate, unrecognized, and unworthy of investment for the immediate present.</p>
<p dir="ltr">In Thaler’s defense, however, there are several issues that I also take with the courts’ responses to his lawsuit; namely the dismissal of AI-generated works on the grounds that randomness plays a role in the creation of the final output. As I mentioned in past articles, any Fluxus artist, performer, painter, and photographer knows that randomness is a fundamental part of the process. The courts’ position, that randomness can only exist in the physical world and digital randomness is somehow illegitimate, is short-sighted. The same goes for arguments over intentionality.</p>
<p dir="ltr">The USCO has also proven to be myopic on other occasions, <a href="https://fingfx.thomsonreuters.com/gfx/legaldocs/klpygnkyrpg/AI%20COPYRIGHT%20decision.pdf">giving copyright protections to parts of a comic book, but not to the images in it</a>, which were generated using Midjourney, in 2023. Although technically feasible, in practice, a comic book is not something you can, or should, dissect. This piecemeal approach betrays the USCO’s genuine lack of understanding of AI-generated works, or perhaps reveals their hesitation in making waves. It is also worth noting that Trump fired the head of the USCO in May and has not appointed a new one since, which must only be making things more complicated in matters that have far-reaching consequences like deciding whether or not AI-generated imagery can be copyrighted.</p>
<p dir="ltr">Much like the rise of NFTs, the rise of AI is too mired in hyper-commercial slop at this juncture to be fully embraced by the public, at least for the immediate present. We will eventually reach an equilibrium and AI-generated art will be part of the landscape. But this lawsuit, this particular lawsuit, has so far done more harm than good. Thaler might be better off regrouping, and having some conversations with us before moving forward with his quest.</p>
<p> </p>
