Beware GenAI for Book Covers
Chasing the bottom line and convenience has gone too far
Hey there! Real quick—I usually only post twice a month: a roundup of my top picks of book covers coming out that month and a wild card newsletter. Just this month, we’ll be doing it the other way ‘round. I can’t sit on this one any longer, so that’s why. Anyways, on with the show!
I’ve been thinking about writing this article on and off since August 2024. Since then, GenAI, its capabilities, whatever version it’s on, how much catch-up businesses have been playing, and how laws around the world have been responding have progressed quite far and viciously so. Most recently, ChatGPT’s newest updates have revealed the ability to generate images in Studio Ghibli director and animator Hayao Miyazaki’s iconic visual style.
In my opinion (but also one echoed across the Internet): it’s just the newest low of GenAI’s copyright-violating, humanity-disrespecting, intellect-degrading progress1.
The discussion around using GenAI, its advantages and limitations, extends far, far beyond the book publishing and art worlds. For relevancy, we’ll just look at its application in book cover designing, but sentiments will likely apply elsewhere.
It is worth keeping in mind through a marketing lens what using GenAI *willingly* (and not imposed by one’s boss) may tell readers and colleagues about prompters and the GenAI work (may it be a generated book cover, an entire book, and otherwise).
What’s the biggest takeaway? You’re cheap.
We are not at a point where using GenAI stock images, even to a tiny portion that is composited into a larger image, is universally acceptable for a whole host of legal, moral, social, and environmental reasons (but whether corps from big to small would comply is another messy matter entirely).
Like any organisation or individual that knowingly and willingly uses GenAI, the primary reasons why publishers, individual designers, and/or authors turn to GenAI are because it’s:
Easy
Free
Fast
Produces objectively good visual assets
In turn, one can save time and money and earn more profit.
Are those logistical, utilitarian viewpoints of GenAI and the benefits that they offer? Sure.
But just because we can use it doesn’t mean we should.
Also, people are resourceful. Readers and colleagues can find out (or see) that GenAI was used.
One of the drawing points of the arts and any tangible outputs associated, including the marketing-intersected product of book covers, is that it involves hard work.
We can empathise, and moralise, with putting hard work into something because, at the very least and to some capacity, it means that the creator cared about what they did and made. Everyone has had to put effort into something. We may even envy those whose efforts have great results, may it be a visually appealing book cover that many compliments or otherwise.
Ideally, all that hard work—the skills utilised, the process, the output—is done to create and in exchange for value.
It’s understandable if one lacks the skills or resources to make, commission, or task employees to make a book cover for oneself, especially a really good one that costs some big bucks.
But using technology that scraped the work of other artists without permission or credit? Where one might credit themselves for the output they prompted? Where the output lacks finesse because of the lack of care beyond the output’s appearance, thereby devaluing the value of their business in the long-term? Where one may spout about how wonderful their pictures are while simultaneously putting down the process of art-making and failing to understand how art can be valued?
At the very least, one is producing and/or distributing work that is not copyrightable and unethical by using genAI.
In the business context, selling work without proper credit (which is nigh impossible with GenAI’s training) to the people the data is taken is highly risky. Even if it technically isn’t because copyright laws are being tested left and right and the volume of output is too fast for courts to catch2 and create new laws to account for, it’s better not to mess around with lawfully grey areas or angry readers. So just because it’s convenient for prompters and corps to create visual assets doesn’t mean it’s fair, sustainable, legal, or cheap in the long run (the assets may not be copyrightable but would cost you in the long run).
In fact…
Everything comes at a cost, may it be money, time, effort (energy and skill), or something else.
Whether we’re talking about creating book covers (plus the text themselves, marketing, printing, all that jazz) from the standpoint of indie authors and presses, trad publishing houses, freelance designers and illustrators, in-house marketers—virtually anyone involved in the process of producing a commercial good—money, time, and effort are some of the bare minimum costs involved.
GenAI, on paper, is intended to cut down expenses, save time, and produce an output that its generator is pleased with for its users. Theoretically, it is a highly promising, implementable tool to increase productivity and lower costs.
But there are costs involved that not even GenAI can reduce.
Publishing such generated book covers can affect one’s reputation, credibility, earning potential and loyalty potential from target audiences.
Word-of-mouth can spread from target readers whose ethics don’t align with the use of GenAI to others, thereby limiting one’s earning potential from target readers, the likeliest biggest earning group, and discrediting oneself from earning their loyalty. Public backlash and decrying from industry colleagues and readers can stain the perception of the book and/or its cover, such as the case with Sarah J Maas’s popular YA fantasy series in 2023 under Bloomsbury Publishing.
Considering the sentiment of GenAI amongst artists, I’d wager that someone who shamelessly uses it out of their own volition, and not solely because of corporate instruction, or a company that endorses it should be, at the very least, approached with wariness.
What we must strive for is a balance between integrity, practicality, and sustainability for the sake of encouraging healthy growth and competition for our practises.
TLDR & Beyond…
Creativity is a skill like any other, a highly useful and pleasurable capability that can be applied in any field, highly sought after, and compensable. When dealing with creatives, interacting with them, being one of them, one better expects to respect the trade.
Businesses operating within the arts, like book publishing, are in choppy waters because of the highly subjective nature of the products they sell, but they can be successful. Book covers are one component that helps get books out there and generate revenue. It’s a business tool, and business does have a strong focus on doing things at a profit. That’s where GenAI comes in and saves the day on cutting down expenses and boosting output, supposedly.
GenAI is, understatedly, undergoing many growing pains. It’s not at a stage, if it can even get to that point, where there’s little environmental, legal, cultural, societal, developmental, operational, and last but not least, ethical, drawbacks from using it as a *tool*. Very few things in this universe can please everyone—there will always be someone who gets displeased—but genAI can’t be lumped into such an elusive category on its “merits” of revolutionising productivity, democratising creativity, and increasing profit margins when it’s so fraught with issues that are headaches for corporations, lawyers, artists, and readers.
Within our current global environment, seeing what else is occurring alongside the genAI development, and observing the different responses towards it, the hype and promises of GenAI are still too early to tell if they will be delivered on.
Yet, I rest my case that the extreme pro-arguments and corporations jumping on the tech train to not be at a disadvantage downplay (or egregiously degrade) the sabotaging of the field, industry, skillset, and perceptions of the arts. As well, they fail to slow down, center, and consider how our very competencies and empathy can be elevated alongside STEM innovations that push humanity to new heights rather than pushed aside and pushing us back.3
Individually, there’s nothing we can do to stop corps or other people from endorsing and incorporating it into their workflows. If you choose to use it, for book covers or otherwise: be aware of what you’re getting into, do your homework, and proceed with caution.
Whew!
That was lengthy, but it needs to be stressed.
Until next time.
https://www.techtarget.com/searchenterpriseai/tip/Generative-AI-ethics-8-biggest-concerns
https://slejournal.springeropen.com/articles/10.1186/s40561-024-00316-7#:~:text=Overreliance%20on%20AI%20dialogue%20systems,potentially%20diminishing%20individual%20judgment%20skills.
https://www.dentons.com/en/insights/newsletters/2025/january/14/dentons-intellectual-property-hub/fakes-made-easy
https://www.reuters.com/practical-law-the-journal/litigation/generative-ai-copyright-2024-10-01/
https://pmc.ncbi.nlm.nih.gov/articles/PMC11239631/