The October 2026 edition of Steam Next Fest started on October 19, allowing users to explore Steam’s massive catalog of games with free playable demos. Millions of PC gamers logged on but on…
It’s about context. I’m working on a game right now, an idea I’ve had for decades but has always been out of reach. AI has helped me a lot to develop it quickly while also working a full time job. And I’m also working hard to avoid disclosing the AI use - not because I don’t want people to know but because of the deserved stigma around AI slop.
Steam’s current stance (last time I checked) is that if the AI makes assets the player sees, hears, reads or otherwise experiences first-hand, then you must disclose it. But if the AI is used as a tool to produce the code that runs the game, it does not require disclosure. I personally agree with this and have tried to adhere to these rules.
In practice, it means that rather than having some generative tool make my in-game textures, I instead had AI create code that procedurally looks at 3D assets and defines how an artist would draw lines to represent that object if they saw it in real life. And when it comes to audio, I’m doing my own Foley work. For all of the dialogue in the game, I’m writing it myself, which is going to amount to hundreds of hours of work.
But it’s worth it to produce a game with a human touch to it, which doesn’t need to disclose use of AI-created assets. If there was a requirement to disclose use of an LLM in producing the backend code, I gladly would. AI has saved me countless hours by writing the game’s source code, which would easily have taken a year or more for me to write and debug by hand. I just don’t want my game to be lumped in with games that are churned out from an AI without an eye for quality, completeness, or a unique look and feel to the assets. I’ve put too much work into this game for someone to take one look at an AI tag and decide that it’s low-quality slop.
So, based on how you describe it, you’re trying to avoid having to make a required disclosure by generating code that makes the visuals rather than generating the visuals directly? To put it in the way I see it: you’re willfully trying to find and abuse a loophole in the disclosure policy because you know having that disclosure makes your game less desirable to a potential audience?
This is why there’s very little trust for anyone who uses genAI. Rather than recognize that its use is creatively bankrupt, ethically questionable, and disrespectful towards the art form it’s emulating, they’ll use any little loophole or technicality to justify deceiving their audience.
Respectfully, I get that you have a worthwhile story to tell and are putting a lot of work into the writing, but either make the game with the help of genAI and make the disclosure, or tell the story in a medium using the skills you have without resorting to genAI. Trying to weasel your way out of making the disclosure is, in my opinion, more disrespectful than the low quality spam.
It’s not a loophole. It is the actual intent of the AI disclosure rule. Generative AI uses inference to produce visuals. I am using AI to create procedural code that draws lines in the way I want them drawn. Using the curves and styles I have defined. You can argue that you want AI usage in backend code to be disclosed as well, which I would probably agree with. But I don’t agree that building logic to define how 2D lines get rendered is the same as prompting an AI for an image and using that image as an asset in a game. They are completely different ways of producing graphics and look entirely different from one another. I do not want to make a game where the art style looks AI-generated. And mine does not look that way, because the art is not made by AI. Still, I respect and appreciate your opinion on the topic and would be interested in hearing further feedback on the process I’m using.
As much as I agree that there is a difference, it still feels like trying to skirt around the disclosure. Clever way to avoid looking generated at least, I’ll give you that. Agree to disagree on this.
I’m very staunchly against using genAI for anything in the games I play, so if I find out a game uses it in any way my opinion of it and its maker quickly shifts negative. If I played a game that doesn’t disclose any genAI use and later find out genAI was used, I will go out of my way to avoid any further games made by the same people and leave a negative review citing the genAI use as the primary reason.
In my actual opinion, creatives using genAI to do work in the mediums they don’t work in is a betrayal of, and voluntary isolation from, people who should be your peers. I would definitely choose to stop associating with someone I know if they tried to utilize genAI as a part of something creative.
Not everyone has that negative an opinion, or even cares if all the code is generated given they’ll never see or know. It’s a hard-line stance coming from a need I feel to oppose genAI regardless of what it’s generating. Generated code is not something that’s particularly enforceable when it comes to disclosure, anyway.
If you’re following the rules as written then I guess you’re welcome to not disclose genAI use and all I can do is cope.
Yeah, totally understandable. Agree that it’s hard if not impossible to enforce a disclosure on generated code. And also agree that what I’m describing is a code-based means to achieve an art style but a means to skirt disclosure rules that would impact me financially. Both things are true.
I’m hoping they will add a middle ground disclosure where I can differentiate my work from pure genAI slop but also give people like yourself a fair heads up.
On the point about separating ourselves from the people we’re supposed to be working with, I both agree and disagree. In my case, I went to college for animation and game design, but found that none of the jobs I could find out of college would pay enough to cover cost of living. I also heard horror stories about eternal crunch time and studios moving between cities frequently to take advantage of tax incentives. So, I moved home with my parents and got a job at the local computer store. My friend got me a job in a call center as a support technician, but it turned out to be more of a site reliability engineering job under support technician pay. Involved a lot of lateral thinking to solve technical problems and I mostly worked and played video games to try to unwind. I wanted to make something that would allow me to change my circumstances and make money as a creative rather than continually burn out being overworked as a technician. I wrote novels, drew animated shorts, wrote scripts, and made games. But nothing ever felt finished enough to show to other people. I’d always find flaws too deep to fix, or discover it was going to take years to make all the assets I needed and gave up because the work seemed so insurmountable before any payoff was possible. I’ve got about fifteen rich detailed game designs from around twenty years of daydreaming, and not much to show for it.
AI has entirely changed my situation. I left my day job and got a job as a DevOps engineer where I basically has AI teach me all the skills I was missing. I left that job and spent a year trying to build a PaaS. Which like all the other passion projects was aimed at generating enough passive income that I could slow down enough in this rat race and start collaborating with other creatives and get to the point where I can pay them to help me make things.
That PaaS has never been profitable, but it did get me noticed and I’m now a senior software architect at a well-known company. For better or worse, my designs have become central to the product and my work hours have been steadily climbing, currently between 70 and 95 hours per week because so many people depend on my work product for direction and I don’t want to let anyone down. I make enough money to have a house to myself now, but don’t have much time to relax in it. What I can do is set my work laptop on one side, to send it off on AI-driven tasks, while having another laptop on the other side that I give instructions while I wait for my work prompt to come back. I’ve resurrected some of my old ideas and now I’m playing games I was beginning to doubt would ever become a reality.
Hopefully, some of these will be profitable enough that I can both support myself and hire creative people to work on more involved projects. I do see your point about AI separating creatives who should be working together. But it also feels like my best odds of having time to collaborate in a healthy creative way is by leveraging AI to make some things on my own first. And not to say I can’t find ways to bring artists in on my projects that I’m working on now.
Despite being made in many different styles and genres, my games tend to have one commonality to them. They introduce some kind of asset submission pipeline where creators can make skins or customizations on top of my games and get paid for their work in a way that fairly balances the time I’ve put in and their submitted work against the benefits our works collectively provide to the player.
Because I am deeply concerned about the direction the world is taking. The ease with which corrupt political parties deceive the populace. The control that the super wealthy have over almost everything. The fact that they will soon be able to do most things without the need for any kind of proletariat. I don’t think the non-wealthy people will get basic income on a platter. They are more likely to be disposed of when they are no longer necessary. It’s already happening.
So, part of my aim is to provide jobs for people in a post-job society. But of course, people without jobs have no money to spend on games. So, the primary purpose of my working insane hours and developing these old game designs is to make money. Specifically, to make enough money that I can leave my job and finish my PaaS. Because that hosting platform will become the foundation of a software creation platform which allows people to democratically create software, in a structure without a competitive hierarchy. One which runs not on capital but collective reasoning between confirmed real humans. And the means of that production as well as the profits of everything produced belong to the ones putting in the work to make the system successful. Those are the rules I need to codify and ensure cannot be corrupted once I allow my system to fork and federate.
And only once we rebuild and improve all of our existing software, currently owned by greedy companies that answer to profit-obsessed shareholders, will the act of living stop funneling profits to the ultra rich. If you can’t change that reality, then it’s only through effective voting and responsible governance that we can regain control of our civilization. And I have stopped holding my breath on that account. These AI companies are spending a lot more than they are making right now because it’s in their best interests to get people hooked and train on the users. It’s only during this window that it’s financially possible for the general populace to create and market tools that will replace the profit-driven platforms of today.
AI does indeed contribute to horrific climate change, enables the ultra rich to disconnect their futures from the existence of the masses, and gets its users addicted in ways that are already hard to reverse. It is nonetheless my opinion that leveraging it heavily in the months and years to come is by far the best shot we have to take the game of the oligarchs and turn it against them. I may be wrong. But you now have all the logic that has brought me to this place.
I can only hope we can find a path that is best for us all.
TL;DR: i think you are correct and most of our differences may be in the perspective from which we view a shared goal
I don’t particularly agree with that stance. In terms of what I’m dealing with in the game, there’s really no difference between the developer asking an AI to make an asset filter and them knowing they need to Google kuwahara image abstraction .
When it comes to making an asset, I care that someone thought about what it should look like. Having it generate the asset is just hitting shuffle until you stop. Having it build the filter or tool is no different than finding a tool or filter and then using it to make the asset.
Making the backend does not make you a good game dev, you also have to make a fking game thats enjoyable to play, sht so many games don’t require coding to make, I see no difference between that and directing an ai on the backend. Or do you hate no code engines meant for ppl to make games easily?
It’s about context. I’m working on a game right now, an idea I’ve had for decades but has always been out of reach. AI has helped me a lot to develop it quickly while also working a full time job. And I’m also working hard to avoid disclosing the AI use - not because I don’t want people to know but because of the deserved stigma around AI slop.
Steam’s current stance (last time I checked) is that if the AI makes assets the player sees, hears, reads or otherwise experiences first-hand, then you must disclose it. But if the AI is used as a tool to produce the code that runs the game, it does not require disclosure. I personally agree with this and have tried to adhere to these rules.
In practice, it means that rather than having some generative tool make my in-game textures, I instead had AI create code that procedurally looks at 3D assets and defines how an artist would draw lines to represent that object if they saw it in real life. And when it comes to audio, I’m doing my own Foley work. For all of the dialogue in the game, I’m writing it myself, which is going to amount to hundreds of hours of work.
But it’s worth it to produce a game with a human touch to it, which doesn’t need to disclose use of AI-created assets. If there was a requirement to disclose use of an LLM in producing the backend code, I gladly would. AI has saved me countless hours by writing the game’s source code, which would easily have taken a year or more for me to write and debug by hand. I just don’t want my game to be lumped in with games that are churned out from an AI without an eye for quality, completeness, or a unique look and feel to the assets. I’ve put too much work into this game for someone to take one look at an AI tag and decide that it’s low-quality slop.
So, based on how you describe it, you’re trying to avoid having to make a required disclosure by generating code that makes the visuals rather than generating the visuals directly? To put it in the way I see it: you’re willfully trying to find and abuse a loophole in the disclosure policy because you know having that disclosure makes your game less desirable to a potential audience?
This is why there’s very little trust for anyone who uses genAI. Rather than recognize that its use is creatively bankrupt, ethically questionable, and disrespectful towards the art form it’s emulating, they’ll use any little loophole or technicality to justify deceiving their audience.
Respectfully, I get that you have a worthwhile story to tell and are putting a lot of work into the writing, but either make the game with the help of genAI and make the disclosure, or tell the story in a medium using the skills you have without resorting to genAI. Trying to weasel your way out of making the disclosure is, in my opinion, more disrespectful than the low quality spam.
Thanks for an intelligent take in a thread full of slop logic.
It’s not a loophole. It is the actual intent of the AI disclosure rule. Generative AI uses inference to produce visuals. I am using AI to create procedural code that draws lines in the way I want them drawn. Using the curves and styles I have defined. You can argue that you want AI usage in backend code to be disclosed as well, which I would probably agree with. But I don’t agree that building logic to define how 2D lines get rendered is the same as prompting an AI for an image and using that image as an asset in a game. They are completely different ways of producing graphics and look entirely different from one another. I do not want to make a game where the art style looks AI-generated. And mine does not look that way, because the art is not made by AI. Still, I respect and appreciate your opinion on the topic and would be interested in hearing further feedback on the process I’m using.
As much as I agree that there is a difference, it still feels like trying to skirt around the disclosure. Clever way to avoid looking generated at least, I’ll give you that. Agree to disagree on this.
I’m very staunchly against using genAI for anything in the games I play, so if I find out a game uses it in any way my opinion of it and its maker quickly shifts negative. If I played a game that doesn’t disclose any genAI use and later find out genAI was used, I will go out of my way to avoid any further games made by the same people and leave a negative review citing the genAI use as the primary reason.
In my actual opinion, creatives using genAI to do work in the mediums they don’t work in is a betrayal of, and voluntary isolation from, people who should be your peers. I would definitely choose to stop associating with someone I know if they tried to utilize genAI as a part of something creative.
Not everyone has that negative an opinion, or even cares if all the code is generated given they’ll never see or know. It’s a hard-line stance coming from a need I feel to oppose genAI regardless of what it’s generating. Generated code is not something that’s particularly enforceable when it comes to disclosure, anyway.
If you’re following the rules as written then I guess you’re welcome to not disclose genAI use and all I can do is cope.
Yeah, totally understandable. Agree that it’s hard if not impossible to enforce a disclosure on generated code. And also agree that what I’m describing is a code-based means to achieve an art style but a means to skirt disclosure rules that would impact me financially. Both things are true.
I’m hoping they will add a middle ground disclosure where I can differentiate my work from pure genAI slop but also give people like yourself a fair heads up.
On the point about separating ourselves from the people we’re supposed to be working with, I both agree and disagree. In my case, I went to college for animation and game design, but found that none of the jobs I could find out of college would pay enough to cover cost of living. I also heard horror stories about eternal crunch time and studios moving between cities frequently to take advantage of tax incentives. So, I moved home with my parents and got a job at the local computer store. My friend got me a job in a call center as a support technician, but it turned out to be more of a site reliability engineering job under support technician pay. Involved a lot of lateral thinking to solve technical problems and I mostly worked and played video games to try to unwind. I wanted to make something that would allow me to change my circumstances and make money as a creative rather than continually burn out being overworked as a technician. I wrote novels, drew animated shorts, wrote scripts, and made games. But nothing ever felt finished enough to show to other people. I’d always find flaws too deep to fix, or discover it was going to take years to make all the assets I needed and gave up because the work seemed so insurmountable before any payoff was possible. I’ve got about fifteen rich detailed game designs from around twenty years of daydreaming, and not much to show for it.
AI has entirely changed my situation. I left my day job and got a job as a DevOps engineer where I basically has AI teach me all the skills I was missing. I left that job and spent a year trying to build a PaaS. Which like all the other passion projects was aimed at generating enough passive income that I could slow down enough in this rat race and start collaborating with other creatives and get to the point where I can pay them to help me make things.
That PaaS has never been profitable, but it did get me noticed and I’m now a senior software architect at a well-known company. For better or worse, my designs have become central to the product and my work hours have been steadily climbing, currently between 70 and 95 hours per week because so many people depend on my work product for direction and I don’t want to let anyone down. I make enough money to have a house to myself now, but don’t have much time to relax in it. What I can do is set my work laptop on one side, to send it off on AI-driven tasks, while having another laptop on the other side that I give instructions while I wait for my work prompt to come back. I’ve resurrected some of my old ideas and now I’m playing games I was beginning to doubt would ever become a reality.
Hopefully, some of these will be profitable enough that I can both support myself and hire creative people to work on more involved projects. I do see your point about AI separating creatives who should be working together. But it also feels like my best odds of having time to collaborate in a healthy creative way is by leveraging AI to make some things on my own first. And not to say I can’t find ways to bring artists in on my projects that I’m working on now.
Despite being made in many different styles and genres, my games tend to have one commonality to them. They introduce some kind of asset submission pipeline where creators can make skins or customizations on top of my games and get paid for their work in a way that fairly balances the time I’ve put in and their submitted work against the benefits our works collectively provide to the player.
Because I am deeply concerned about the direction the world is taking. The ease with which corrupt political parties deceive the populace. The control that the super wealthy have over almost everything. The fact that they will soon be able to do most things without the need for any kind of proletariat. I don’t think the non-wealthy people will get basic income on a platter. They are more likely to be disposed of when they are no longer necessary. It’s already happening.
So, part of my aim is to provide jobs for people in a post-job society. But of course, people without jobs have no money to spend on games. So, the primary purpose of my working insane hours and developing these old game designs is to make money. Specifically, to make enough money that I can leave my job and finish my PaaS. Because that hosting platform will become the foundation of a software creation platform which allows people to democratically create software, in a structure without a competitive hierarchy. One which runs not on capital but collective reasoning between confirmed real humans. And the means of that production as well as the profits of everything produced belong to the ones putting in the work to make the system successful. Those are the rules I need to codify and ensure cannot be corrupted once I allow my system to fork and federate.
And only once we rebuild and improve all of our existing software, currently owned by greedy companies that answer to profit-obsessed shareholders, will the act of living stop funneling profits to the ultra rich. If you can’t change that reality, then it’s only through effective voting and responsible governance that we can regain control of our civilization. And I have stopped holding my breath on that account. These AI companies are spending a lot more than they are making right now because it’s in their best interests to get people hooked and train on the users. It’s only during this window that it’s financially possible for the general populace to create and market tools that will replace the profit-driven platforms of today.
AI does indeed contribute to horrific climate change, enables the ultra rich to disconnect their futures from the existence of the masses, and gets its users addicted in ways that are already hard to reverse. It is nonetheless my opinion that leveraging it heavily in the months and years to come is by far the best shot we have to take the game of the oligarchs and turn it against them. I may be wrong. But you now have all the logic that has brought me to this place.
I can only hope we can find a path that is best for us all.
TL;DR: i think you are correct and most of our differences may be in the perspective from which we view a shared goal
I don’t particularly agree with that stance. In terms of what I’m dealing with in the game, there’s really no difference between the developer asking an AI to make an asset filter and them knowing they need to Google kuwahara image abstraction .
When it comes to making an asset, I care that someone thought about what it should look like. Having it generate the asset is just hitting shuffle until you stop. Having it build the filter or tool is no different than finding a tool or filter and then using it to make the asset.
Making the backend does not make you a good game dev, you also have to make a fking game thats enjoyable to play, sht so many games don’t require coding to make, I see no difference between that and directing an ai on the backend. Or do you hate no code engines meant for ppl to make games easily?