In Theory: AI Art In Board Games

While I generally try to avoid controversy on this site and stick to what I enjoy most, reviewing games, talking about games and, well… more games, now and then a subject comes along that is simply impossible to ignore.

This is one of those subjects.

As someone who reviews board games, I’m increasingly running into games that use AI-generated artwork, which means I have to make a decision about how I’m going to treat them. Do I ignore it? Mention it? Penalize it? Celebrate it? Pretend it isn’t there and hope nobody notices?

Sooner or later, I have to put my cards on the table, explain my position and live with the consequences. The internet being what it is, remaining silent makes you complicit, while saying anything at all guarantees that one tribe or the other will decide you’re the villain of the week. It’s a remarkable system we’ve built for ourselves in which you are always left with a lose-lose scenario.

I’ve touched on AI artwork in a few previous reviews, most recently Syncanite Foundation and Kingdom Legacy, and those conversations have helped me work out where I stand. But scattered comments buried inside reviews aren’t enough anymore. I need to make dealing with AI art work part of my rating system, so that I can respond to it in an objective and fair way.

So this article will be the official GamersDungeon position on AI artwork in board games, how I’m going to approach it as a reviewer and, most importantly, how it will impact the rating/scores that games receive going forward.

The Controversy

Unless you’ve been hiding under a particularly large and comfortable rock, you’re probably aware that AI is everywhere. In the tabletop hobby, and especially in board games, role playing games and miniature games, AI-generated artwork has become one of the most divisive subjects around.

Mention it in a comment section, and you’ll usually have enough material for a three-day flame war.

While there are dozens of individual arguments and plenty of grey areas, the debate generally revolves around three major points.

The first is that generative AI art is fundamentally a form of theft. The argument is that AI models are trained on existing artwork created by real artists and then produce derivative images without permission, attribution or compensation. In other words, the machine is standing on the shoulders of artists who never agreed to hold it up.

Adobe Firefly combats Generative AI theft by training it’s AI on public domain images and images willingly provided by artists. This is just one of many unique methods that put to question the argument that Generative AI images are theft. Hence the problem with this argument.

The second argument is economic. Every AI image used by a publisher is potentially one less commission for a human artist. If a company can generate an illustration in minutes instead of paying an illustrator, more profit stays with the publisher, while fewer opportunities exist for the people who built the artistic foundation AI relies upon. To critics, it isn’t just replacing jobs, it’s replacing them with something built from the work of those same artists.

Finally, there is the quality argument. Critics often describe AI art as soulless, repetitive, and creatively hollow, produced by systems that consume enormous amounts of computing power and energy simply to flood the internet with an endless stream of technically competent but artistically disposable images. The term AI slop didn’t appear out of nowhere.

There are plenty of smaller arguments, edge cases and philosophical rabbit holes that could fill an entire series of articles, but these three points are the heavy hitters. If I can explain where I stand on them, then I can also explain how AI artwork will be treated in reviews here on GamersDungeon going forward.

First, however, we have to talk about the elephant in the room.

Circumstances Matter

I’ve never had much patience for ivory tower thinking or the modern habit of treating every issue like it’s a football match where you have to pick a side and spend the next six months screaming at the other team.

The real world is a lot messier than that.

Real people have real jobs, real businesses, real families and real bills to pay. Artists, publishers, designers, consumers and even the people building AI tools all have different incentives and different circumstances. Any position that completely ignores one side in favour of ideological purity is, in my opinion, more interested in winning an argument than solving a problem.

Kingdom Legacy and Fryxelius Games is a great example of circumstances mattering. This is a family run business of creative people who are doing their best to bring great games to us. They however like all businesses have to make compromises. In the case of Kingdom Legacy, your talking about producing art for 140 quardruple sided cards requiring around 500 images for a game that can’t cost more than 10-15 bucks for it to be marketable. Had Fryxelius games hired an artist to create these images this game would never see the light of day and if it did it would cost more than anyone would be willing to pay for a game that is effectively a box with 140 cards in it.

That isn’t particularly useful to me.

So I’m not going to approach AI artwork from the perspective of absolute morality, nor am I going to pretend that technological progress can simply be wished away. My position has to account for the many people affected by it, which means it’s inevitably going to be a compromise.

To put it plainly, I’m not taking the easy route of saying “I refuse to review games with AI art” and I’m equally not going to shrug and say “I don’t care, embrace the future.”

Somewhere between those two extremes is a position that I think is both fair and practical. Whether you ultimately agree with it or not, I think it’s worth explaining how I arrived there before I tell you what the policy will be.

AI Art is Stealing

This is probably the biggest argument against AI-generated art, and it’s also the one I find the hardest to apply in practice.

Not because I know it isn’t true, but because I don’t know that it is.

I’m not an AI engineer, and I’m certainly not qualified to explain exactly what every image model is doing behind the scenes. More importantly, not every AI is trained the same way. Some models are trained on enormous collections of scraped images, while others are built from artwork that has been voluntarily submitted or properly licensed by the artists involved.

Those are very different situations.

A good example is Kingdom Legacy. After doing some research for that review, I discovered that the publisher uses an AI trained on artwork freely contributed by artists. If that’s the case, then the blanket statement that “AI art is theft” simply doesn’t apply.

The problem is that I can’t realistically investigate the AI training methods behind every game that uses AI-art I review, and even if I tried, publishers have no obligation to explain their workflow or be completely transparent about it.

So what am I supposed to do? Assume everyone is guilty until proven innocent? Or assume everyone is acting ethically until proven otherwise?

Neither approach seems particularly reasonable.

For that reason, I can’t base my review policy on the argument that AI art is inherently stealing. There are simply too many variables, too many different models and too many different ways of using the technology for me to conclude that every instance of generative AI is automatically unethical.

That’s not the same as saying the concern isn’t valid. It’s saying that, as a reviewer sitting behind a keyboard trying to decide whether a board game deserves a 3.5 or a 4, I don’t have enough information to make that judgment consistently or fairly.

So, for the most part, I set this argument aside. Not because I dismissed it, but because I don’t think it provides a practical foundation for a review policy.

They Took’ma’job!

I’m going to keep this one relatively short. Technology replaces people. It always has.

The printing press replaced scribes, photography replaced portrait painters, tractors replaced farm workers, digital distribution replaced video rental stores and the internet made life very uncomfortable for anyone who thought selling encyclopedias door to door was a long-term career plan.

We can resist it, protest it and argue about whether it’s a good thing, and sometimes those arguments are completely justified. History, however, has a habit of continuing anyway.

Dragonfoot Forums, one of the oldest D&D forums in existance has recently taken the decision to ban AI art from their forums and will moderate AI created material published through their site. This sort of reaction to AI art is common. Gamers everywhere are rejecting AI normalization and for good reason. Art is culture and AI is erasing it.

My personal philosophy has always been simple. Adapt and survive. Do I think it’s a good thing if artists lose work to AI? Absolutely not. But that isn’t actually what influences my reviews.

What influences my reviews is that I have yet to see AI-generated artwork that was worth replacing a human artist for in the first place.

That’s the important distinction.

I’m not making a moral judgment about technological progress. I’m making an artistic judgment about the end result.

And, quite frankly, I’m not impressed.

To me, AI artwork is shallow, repetitive and creatively uninteresting. I have no desire to sit here debating whether a particular image is “good AI” or “bad AI” any more than I want to debate whether instant coffee is “good coffee.” At best, it’s mediocre. At worst, it’s visual wallpaper that exists solely because someone needed a dragon by Tuesday afternoon.

Talent is something people develop over years of practice. Style is something people earn through experience, experimentation and failure. If the artwork in a game can be produced by me, my neighbour and a reasonably motivated golden retriever typing prompts into the same generator, then I struggle to assign much artistic value to it.

As a reviewer, that matters.

If I believe components contribute to the overall experience of a board game, then artwork is part of that equation, and artwork that I consider generic, uninspired or interchangeable should naturally be reflected in the score.

But even that isn’t really the heart of the issue.

The real reason AI art matters to me is something much more fundamental.

AI Art Has No Soul

This is the argument that ultimately matters to me.

I’ve already said that I’m unconvinced by the blanket claim that all AI art is theft and equally unconvinced that I can somehow stop technological progress by refusing to acknowledge it.

None of that changes the simple fact that I don’t like AI art. Not a little. Not “when it’s used badly.” I don’t like it at all.

The ecological cost, the enormous computing resources and the economic disruption only reinforce an opinion I already have, which is that the end result simply isn’t worth it. It’s an extraordinary amount of effort and energy being spent to produce something that, in my eyes, is artistically mediocre.

To me there is no masterpiece hiding inside AI-generated artwork, only different flavours of competent wallpaper. It can be technically impressive, visually striking and even useful, but I have yet to see anything that makes me stop and appreciate the person behind it.

Because there isn’t one.

Syncanite Foundation is one of those rare exceptions where I thought the AI art was well curated. It was the first review I ever did for a project with AI art however and I wasn’t sure what to do with it. In the end, I chose to just judge the art as I would any other, but it felt wrong. I don’t want to judge AI art, it felt empty, like I was speaking to a void rather than complimenting a persons hard work. This is what I want to avoid having to do in my reviews.

What makes art meaningful to me isn’t perfection. It’s the evidence that another human being sat down with a skill they spent years developing and created something that could only exist because they chose to make it. The mistakes, the style, the personality and even the imperfections are part of the experience.

That’s the soul.

AI removes the very thing I value most about art and replaces it with automation. It turns creativity into manufacturing, and while that may be efficient, efficiency has never been the quality that made me love board games, role playing games or miniature games in the first place.

So this is where I draw my line. Not because I think AI should be banned. Not because I think everyone who uses it is acting unethically. And not because I believe technology can be put back into the bottle.

But because, as a reviewer, I want to reward human creativity wherever I find it. Choosing a human artist over a prompt is, in my opinion, an investment in the very creative spirit that makes this hobby worth celebrating.

That’s my protest.

To me, replacing genuine artistic expression with AI artwork is like spray painting over a beautiful mural. The person holding the can may have perfectly reasonable motivations and the paint may even look neat from a distance, but something uniquely human has still been covered up in the process.

And that, more than any legal or economic argument, is why AI artwork will matter in my reviews.

Conclusion

I should probably end with a confession. I use AI art in my own projects.

When I wrote my D&D adventure The Lost Citadel, a project I’m genuinely proud of, I used AI-generated artwork for one very simple reason. I couldn’t afford to hire an artist, or perhaps more accurately, I didn’t want to afford hiring an artist. It was a hobby project, I did it for fun, not as a business venture.

That doesn’t suddenly make the artwork great.

If anything, I fully accept that the book is artistically less than it could have been. The illustrations do their job, but they don’t define the identity of the book the way a human artist could have. They lack personality, style and, for want of a better word, soul.

And if someone looked at The Lost Citadel, decided it was AI slop and chose not to buy it, I wouldn’t hold it against them for a second.

I understand the position because I understand the compromise I made.

As a reviewer, however, I don’t think the answer is to draw a line so extreme that any game containing AI artwork is immediately dismissed as worthless.

A board game is more than its illustrations.

It is mechanics, design, theme, writing, balance, playtesting, production and countless hours of work by real people who may have chosen AI art for reasons ranging from budget constraints to simple practicality. Just as I don’t want my own work dismissed solely because I couldn’t afford an illustrator, I’m not going to do that to someone else.

But I also think there should be a clear acknowledgement that AI artwork is not something I value as an artistic contribution.

So this is the new policy at GamersDungeon.

Any game that uses AI-generated artwork will receive a maximum of 1 star in the Theme category of my reviews.

That doesn’t mean the game is bad. It doesn’t mean I won’t recommend it. It doesn’t mean the designers are lazy or unethical.

It simply means that, in my view, AI-generated artwork does not meaningfully contribute to the artistic identity of a board game and therefore cannot receive a higher score in a category where artistic presentation is a major consideration.

Everything else will still be judged on its own merits. Great mechanics will still be great mechanics. Brilliant design will still be brilliant design. An exceptional game can still receive an exceptional overall score.

In fact, Kingdom Legacy: Exploration managed a respectable 3.15 out of 5 despite receiving only 1 star for Theme.

So this isn’t a boycott. It’s a statement of values.

If you choose AI artwork instead of human artistry, I’m not going to refuse to review your game, and I’m not going to pretend the rest of your work doesn’t matter.

Gamersdungeon.net rating system will be updated with the new AI based rule put into effect. For me, this is a compromise and the most appropriate way to handle AI. It may change in the future, but for now I feel like it’s good middle ground I can work with.

I’m simply going to score the art exactly as I see it. The absence of effort, the equivillant of copy/pasting it from some other source, a non-contributor.

And from this day forward, that’s how AI-generated artwork will be handled on GamersDungeon.net.

And that’s all, folks!

Leave a Reply

Your email address will not be published. Required fields are marked *