Substance not Style: What Generative AI Steals from Artists and the World

I initially planned a different article, but this development seems much more important. Judge Vince Chhabria’s recent remarks on the Kadrey vs. Meta case should give artists everywhere hope. According to Reuters’ Blake Brittain and Wired’s Kate Nibbs, Chhabria stated:

You have companies using copyright-protected material to create a product that is capable of producing an infinite number of competing products. If you are dramatically changing, you might even say obliterating, the market for that person’s work, and you’re saying that you don’t even have to pay a license to that person—I just don’t understand how that can be fair use.

Nibbs writes that although Meta’s lawyer argued that claims that generative AI harms artists financially are “just speculation,” Chhabria pondered how this might impact the “‘next Taylor Swift’” or “a ‘relatively unknown artist’ whose work was ingested by Meta . . . if the model produced ‘a billion songs in their style.’” In Nibbs’ estimation, “At times, it sounded like the case was the authors’ to lose, with Chhabria noting that Meta was ‘destined to fail’ if the plaintiffs could prove that Meta’s tools created similar works that cratered how much money they could make from their work.” The caveat is that it seems Chhabria believes it will be difficult for the plaintiffs to do this. The onus is now on the plaintiffs, and the creative community at large, to show how generative AI is destroying our livelihoods.

As a songwriter who has not released new work since learning about generative AI, the judge’s remarks felt like an acknowledgment of the harm being done to artists in all disciplines. As I wrote before, generative AI can only give us Beethoven 2.0 and not Beethoven IV or even Beethoven VIII. Beethoven 2.0 sounds so much like Beethoven that many might confuse the digital clone with the original. Beethoven IV and Beethoven VIII are only comparable to the master in that each of these musicians is as original as the original and, therefore, has a similar outsized impact on the musical world. As Chhabria deduced, the fight against allowing generative AI companies to steal from artists in order to train AI that competes with them in their markets is, at its core, about protecting small, emerging, and, by and large, independent artists who may never be able to fulfill their life purposes or, for those who are spiritually inclined, are being dissuaded by the flood of derivative AI “slop” from pursuing their callings and thereby adding the most value they can to the world.

According to the following articles, artists are already losing work or jobs to generative AI:

• “AI Is Already Taking Jobs in the Video Game Industry” by Brian Merchant
• “AI is replacing artists, and here’s the proof” by Paul Hatton
• “Illustrator Kelly McKernan reveals the raw impact of AI on artists’ lives” by Ian Dean
• “AI Reset: Layoffs, RTO, and the New Realities of Work” by Jason Snyder
• “2 years into Unity's long downward spiral, even more employees are being laid off as CEO says it's still 'stretched across too many products'” by Andy Chalk

Consider, also, that researchers from three Canadian universities reported in February 2024 that there were massive layoffs in the video game industry despite companies in that space making huge profits. Moreover, a study by CVL Economics, which was commissioned by groups (including the Human Artistry Campaign) that are fighting AI companies that exploit artists, found that 75% of C-suite executives in the entertainment industry “indicated that GenAI tools, software, and/or models had supported the elimination, reduction, or consolidation of jobs in their business division.” While they also indicated new jobs were being created because of AI, “whether these new jobs will offset inevitable job losses is unclear.”

In addition to losing careers (if they are commercial artists) or day jobs (if they are independent artists) in the creative sector, as well as in sectors like copywriting, artists are also holding off on or not releasing new work or feel compelled to not pursue the arts academically and/or as a profession because of the rampant data theft generative AI companies are engaged in.

Like piracy, generative AI most harms entry-level or emerging artists and thereby forestalls progress in and drains the dynamism of both the arts and the digital world known as the internet, leaving society stuck with increasingly monotonous art and producing a less intelligent and less creative society. This isn’t hyperbole. In addition to a joint study conducted in 2021 by University of Minnesota and University of Cologne researchers, a study conducted this year (in 2025) by none other than Microsoft itself and in collaboration with Carnegie Mellon University further supports the conclusion that generative AI makes users less intelligent and creative.

Meanwhile, as the AI Studio Ghibli and Barbie Box trends show, otherwise unprofitable OpenAI is literally, capitalizing on existing and widely beloved intellectual property in order to both make and market its wares. It appears that without the ability to mimic or create derivative work that relies heavily on existing IP or beloved brands, most people just aren’t interested in generative AI art. This lack of widespread interest without the ability to exploit specific established artists or brands, such as Studio Ghibli, the Barbie franchise, the Simpsons franchise, Sarah Silverman, and Junot Diaz, or, more worryingly, without the ability to exploit emerging artists who are in the process of becoming renowned for what many call style but is more aptly defined as substance (see below) is also why signatures on numerous petitions by artists pushing for protection from generative AI corporations far outnumber signatures on petitions advocating for companies to be able to use copyrighted data to train generative AI without the consent of copyright owners. Here’s the most recent petition against the unlicensed use of copyrighted work for AI training and which has over 50,000 signatures from both major (world-renowned) artists and smaller artists in all disciplines, and here’s an earlier petition with the same aim and which garnered over 3,000 signatures. Compare that to Creative Commons’ petition, organized (it was later revealed) by former Google policy director Derek Slater, which has fewer than 200 signatures and was signed by people who identify as AI artists but mostly use pseudonyms, further undermining the validity of the petition. Why the huge discrepancy? As Variety’s “Generative AI & Licensing: A Special Report,” which was published in October 2024, showed, most human beings understand that what AI companies have done is abhorrent and, therefore, strongly, on questions of consent and compensation, support artists at rates of 77% and 75%, respectively. This report also shows that 76% of the American public believes “there should be restrictions on the ability of AI tools to copy the style of a specific creative artist/writer/musician/creator.” These numbers represent American public perceptions. However, public support for artists’ copyrights persists (and is even stronger) beyond the United States.

The fact that generative AI corporations depend so deeply on intellectual property they do not own to both make and market generative AI models points to a deeper reality: What generative AI steals from artists is substance, which is tied to artistic voice, and not just style, which is superficial. Generative AI takes that which made Hayao Miyazaki’s or Studio Ghibli’s artistic voice so unique that, as recently as a year ago and despite being aware of other artists who influenced Miyazaki, fans, including an art student, couldn’t pin down what made it so particular or different, with the most incisive commenter noting that it has “become a category unto itself.” It is that particularity or artistic voice/substance—that which makes a great artist different and recognizable and so goes beyond mere style—that enables Swifties to distinguish a Taylor Swift song from the works of every other popular Americana-influenced female pop musician (e.g., Sabrina Carpenter or Olivia Rodrigo); that which distinguishes Kanda Bongo Man from Franco; Radiohead from Muse and every other Radiohead-influenced band; Arvo Pärt from Palestrina; Chopin from Ravel; or, more importantly, Nina Simone’s version of “Feeling Good” from Lauryn Hill’s cover of Simone’s cover and both Simone’s and Hill’s versions from the original rendition (which was sung by Gilbert Price and written by Anthony Newley); or, finally and to drive the point further home, even a Hilary Hahn performance of Dvorak’s Violin Concerto in A Minor from a performance of the very same piece by Joshua Bell. Generative AI steals the signature (one-of-a-kind) voice of every great artist without the consent of that artist. This is why I often think of generative AI as Sylar from the NBC TV show Heroes.

In the visual arts (and as anyone who has ever watched Antiques Roadshow knows), the authenticity of a work holds great weight and whole industries are dedicated to provenance. This is also true, to a certain extent, in the literary arts where not just copyright infringement but the lesser offense of plagiarism is frowned upon and results in canceled book deals (e.g., Kaavya Viswanathan). However, generative AI is a master forger. It steals the particular voice, the artistic DNA “or soul”—the substance and not merely the style—of every great creative, thereby devaluing or diverting away from that artist the value of their creativity and labor. In the process, it does something which victims of exploitation understand far better than those who knowingly exploit them: it dehumanizes these artists by siphoning that which makes them and their contributions to our world as unique and valuable as those of any other human in any other field, blurring out both who these artists and those who impersonate them actually are into a ubiquitous uniformity we’ve come to know as “slop,” which is the very antithesis of what copyright seeks to facilitate and cultivate. For many artists, such exploitation kills the spirit.

I hope Chhabria, other judges, and legislators everywhere will recognize how deeply unfair/unjust/unethical, demoralizing, and (I believe) illegal this is and protect both today’s and tomorrow’s art and artists: the human and still unknown Beethovens, Swifts, Simones, Miyazakis, Sargents, O’Keefes, Adams, Parks, Rowlings, Márquezes, Allendes, Millays, Hurstons, Wrights, Bradburys, Hitchcocks, Kurosawas, Van Peebles, Andersons, and more whose works may remain largely unheard, unseen, unread or, worse, unrealized if this injustice continues unchecked.