TLDR: If your game uses generative AI, please disclose this properly. Indiecator doesn't cover games that make use of generative AI.
I didn’t think I would have to keep track of the number, but I was just curious how many rejection e-mails á la “Hey, I don’t cover games featuring generative AI on Indiecator” I have sent out and including three titles today, that makes a hundred. I denied coverage for a total of 100 developers who reached out to me over the past couple of weeks because their games make use of generative AI in some capacity.
Hence, today, I wanted to vent a little bit about this issue… and I also wanted to give some reasoning behind my dislike for the technology and why I’ve got such a firm stance on generative AI in game dev.
The Generative AI Question
For starters, generative AI is used in many capacities in game dev, from “vibe coding” (having an LLM generate code for you) to using Stable Diffusion, LLMs or other tools to generate in-game assets or promotional artwork to using AI to generate “voice acting”, “music” and sound effects.
Generative AI, LLMs and other such tools are, at their core, technologies trained on stolen datasets from real people who never consented to their work being used like this. That includes artists, (voice) actors, musicians, writers (like me), animators, and developers. Now, some companies have claimed that they have “no control” over who they scrape from… and those companies are wrong, as leaks have proven time and time again. On that note, I do recommend this Geekenders episode featuring RubberRoss. He speaks a lot from his experience on the topic, as he (and other creators in the space) are on the Midjourney list, and I do highly suggest giving that a watch!
One of my problems with generative AI is that it makes use of these copyrighted materials and many creatives’ work and effort and then just jumbles it all together into something new that is void of copyright. Money laundering is when you take “dirty money” and then act as if there is a business that it comes from to “clean” the money so that the authorities don’t catch on to the illegal operation you run.
To me, generative AI feels like a form of copyright laundering. A machine doesn’t own copyright, as courts have ruled (Reuters) last year. Purely AI-generated material lacks copyright protection because copyright requires sufficient human authorship. To put it simply, human-made, copyrighted work is fed into a machine that then produces something that may not iself qualify for copyright protection because it lacks sufficient human authorship. The original creators may have never consented to their work being used in the process in the first place, and worse yet, even if they did, they’ll never get properly credited or compensated for the work they put in. The fact that Midjourney’s higher-ups claim to know nothing when they do is evidence of that.
There aren’t enough regulations in place for this kind of stuff, and if the people whose data this technology was trained on consented to the scraping, it would be a different story. Just because they publish it on the internet, that doesn’t mean that consent doesn’t matter. Just because a person leaves the house, that doesn’t mean that they can’t consent anymore to things happening to them.
Beyond that, it’s also a question of job security. Animators in Japan have terrible working conditions, and if they wanted to organise a strike, they’d have less leverage over big production committees because of the existence of AI-generated animation. Similarly, many people can’t work “normal jobs” or make a living off of them due to disabilities, for instance, and many who rely on art to make a living might get replaced entirely by the cheap (if not free) generative AI that people pump out with just a few prompts.
Furthermore, there’s also the environmental component. There is also the component of it effectively having vast and unknown, possibly irreversible effects on our youth and elderly alike. On top of that, generative AI is used to spread fake news, deepfake people, and create pornography of real people who, again, are not consenting to it.
All in all, generative AI may be able to help people as a tool, especially in terms of accessibility, such as AI-powered translation, screen readers, speech-to-text and dictation software, etc., but as it stands, the technology is being used for many things that are not good. One might argue that a tool is just a tool and the people who use it for nefarious things are the issue, but even so, I find that encouraging more people to use it would be wrong for me.
On Reviews
TLDR: I aim to showcase human creativity and craftsmanship in indie games and I refuse to cover games that take the easy way out. I'd rather see MS Paint placeholder art than whatever a soulless machine pumps out.
To me, indie development is all about creativity and thinking outside of the box, and I believe generative AI runs counter to that spirit.
For me, a game loses a significant amount of its charm when I discover that generative AI was used in its creation. I recently covered an incremental/roguelike game that had sound effects entirely consisting of home-recorded sounds, from bubble pops to lighter flicking. I love that. I also have been enjoying Slay the Spire 2 a whole bunch, even if or especially because new events and cards feature beta artwork that was poorly drawn in Paint. It’s exciting to see the game take shape, and I absolutely adore the beta art for the cards!
If you need art, plenty of artists offer their services at a huge range of price points, and there are plenty of ways to adapt your game’s visual direction to your budget. If you cannot afford a particular style, make something different. Work within your limitations. Some of the most interesting indie games ever made exist precisely because their developers embraced what they could create instead of trying to imitate something bigger.
If you cannot afford something, adapt your creative vision to what you can actually make.
If you need to do some programming, there are plenty of tutorials out there, free resources, and lots of documentation. If AI creates subpar code for you or if it hallucinates something, and as a result you end up with frame issues or random bugs, then how are you going to fix them if you don’t understand what is going on under the hood? I have been working on a game myself recently and while I have no clue about programming, I do enjoy discovering new ways of doing things by browsing forums and seeing what Godot can do. I love the sense of discovery whenever I find a solution to some problem I created for myself, and while it may take longer, I find that it does expand my horizons in a meaningful way.
Why do we have all the knowledge and resources of the world at the tips of our fingers if we can’t use them ourselves? Relying on AI for the sake of creating means relinquishing the very thing that makes art “art”. It’s the soul, passion, and heartblood that you rob your game of.
To me, it can turn a game into a product rather than a creative work: Something assembled as quickly and efficiently as possible rather than something a person actually invested their skills, time, and creativity into.
Six out of a hundred developers who reached out to me used generative AI for the writing, translation, artwork, programming, music, and sound effects. So… what did they actually do themselves? I don’t see how they can be credited for any of that work if they didn’t do it, and I see even less why I should play a game that costs 15 bucks and promote it on here when these people are just lining their pockets with the work of other people who never consented to working on this game.
Again, creativity and humanity are what matter to me the most, and I can’t be assured that it’s in the game if the title in question utilises a machine meant to replace the very humanity that makes indie “indie”. Indie, to me, is not just a business classification. It’s about independence of creative expression. Generative AI cuts into that way too much for me.
What annoys me the most is that more than half of the developers don’t disclose it in their e-mails. So, I have to then go to the Steam page, scroll all the way down, get disappointed, reply to the developer, and feel a little bit annoyed. Per email, that’s not a lot of labour… but it adds up.
I’ve not started this passion project of mine to become a sleuth. I started this to highlight great games made by real people and to write articles because I do enjoy talking about things I like.
I love indie games. I love creativity. I love the “thinking outside the box” type games that have come out over the years, and I’d love to see more of that.
Up until this point, I’ve tried to be as polite as possible with developers but I’ve not kept it a secret whatsoever that I don’t cover games with generative AI. I’m very vocal about the fact that I am not a fan of the tech, and in the future, I may just have to go “RTFM” in reply to this sort of stuff, linking to this very post instead of manually typing out a polite reply.
This post was originally published on Indiecator.
If you wanna get in touch with me, feel free to stop by for one of my Twitch streams. For business inquiries, please reach out via email.

I don’t entirely agree with your position, but I respect it for its clarity.
AI needs to be disclosed clearly to make a proper judgement on a work of creativity. There can still be creativity in assembly and origin of concept in something that is largely AI generated, but it is a different beast than something wholly created by human hands.
I have never developed a game, but I have developed a lot of (enterprise) software. I used google to find answers in forums before AI was a thing and that was once looked upon quite negatively. But I found answers more efficiently that way, and was able to adapt them to my own creations. This is I suppose akin to using pre-created animation and sound libraries that can be purchased for various game engines.
I could see AI being something like that: a stepping stone to a more complete creative artifact assembled by a human. If I ever created a game it would be an experiment and I would probably use AI for at least some of the work: maybe 30% of the code logic and 80% of the visual/audio work. I would hopefully remember not to submit it to you for review, however ;)
LikeLike