Public Steam disclosure research25,376 disclosures tracked
Where do games disclose AI use?
Search developers’ Steam disclosures, compare classified AI use across
games, and follow recorded changes.
This index adds cross-game classifications and statistics to
statements published on individual store pages.
Groups reflect automatic analysis of public disclosures; mature records are excluded. These counts do not measure AI use across all of Steam.
Missing or unclear statements cannot establish whether a game uses AI.
The current soundtrack and its album-cover artwork were generated with AI music and image tools and selected by the developer. I am looking to commission original music from human artists; if you make music and would like to hear your tracks in BrickAmp, get in touch at team@brickamp.com. Some level layouts and visual effects were drafted with AI-assisted coding tools and hand-reviewed. All of this content is pre-generated and ships with the game. Nothing is generated live during play, and the game does not connect to any AI service.
The game's code, which procedurally generates all 3D models, textures, sound effects and music at runtime, was written with the help of an AI coding assistant, as were the store description, marketing images and trailer. No AI generation happens while the game is running.
AI was used to assist with coding and to create some textures, which were later reworked by the developer.
The game does not generate AI content in real time
AI tools were used to assist in creating selected promotional store assets and game assets. These materials were or will be later edited, refined, and finalized by human creators.
HELIOBANE was made with the help of generative AI tools.
- Assets: AI was used to generate some of the game's assets, including artwork, voice lines, sound effects, in-game text and translations. Artwork was edited and fitted to the game's 32-colour palette and style before use.
- Music: the soundtrack was made with a custom tracker built for agentic work, by an AI trained on a curated collection of public-domain tracker (MOD) music.
- Some store page and marketing art is also AI-assisted.
Nothing is generated while you play. All content is fixed in the build, and the game does not connect to any AI service.
This game uses a mix of human and generative AI tools. AI was used in some visual art assets, code generation, and portions of the store page text. All writing in the game, voice acting, and music are fully human-created.
Dicebound uses generative AI as part of the development process for some pre-generated content. This includes certain visual assets and promotional/store materials, as well as some drafting support for interface text and localization. All AI-assisted content has been reviewed, edited, and curated by the developer before being included in the game or on the store page. Dicebound does not use live generative AI during gameplay, and players cannot generate content through AI in-game.
Gród is designed, built and tested by a solo developer. Every rule, system and pixel of the game follows my own vision.
AI tools helped with parts of the production: the voice lines, the music and some store artwork were created with generative AI, then selected, edited and fitted into the game by me. The in-game pixel art is drawn by the game's own code, with no AI image generators.
Screenshots and GIFs show real gameplay. The game does not use generative AI while you play.
This game uses pre-generated AI-assisted visual content. Generative AI tools were used during development to help create and refine some 2D artwork and Steam store materials, including Chinese ink-style backgrounds, capsule images, UI/marketing artwork, icons, and related visual elements. All AI-assisted assets were reviewed and edited by the developer before being included. The game does not use live generative AI during gameplay, and no player input is sent to an AI service to generate content.
The game engine is written with substantial AI assistance.
Some dialogues and quest texts are AI generated, and are successively replaced by human-written dialogue. Every piece of flavor text has a label that states the author (human or ai).
The 3D character models are currently generated with AI, and then transformed through a custom pipeline to work with the text-based graphics and CPU rendering. It is possible to replace them with manually crafted models.
Some background environment props and their textures (signage, lamps,
surfaces, decals and small set-dressing objects) were generated with AI
assistance, using AI written procedural code rather than an image or mesh
generator. They ship with the game as scenery.
Everything else is the developer's own work: all writing and in-game text,
game design, game code, audio and sound design, the visual identity, and all
store page art.
No AI generated content is produced while the game is running.
DEADLUCK V was created using generative AI, including its artwork, music, sound effects and written content. This content was generated during development. The game does not generate AI content or connect to AI services during gameplay.
This game includes assets created with generative AI during development (official music, song artwork, and drum animations). Nothing is generated while you play, and the game does not connect to an AI service at runtime. Players cannot send prompts to any AI.
Some store and library capsule art was produced with image-generation tools
(SDXL and a chat assistant) using our own in-game screenshots as the input
image. The underlying game content -- models, textures, code, text -- is
authored by us without generative AI. All generated art was reviewed by us
before publication.
AI models have been used to validate, optimize, rank and test the puzzles and the simulation engine. Soundtrack is procedurally generated music using the same engine that simulates the electronics, powered by the C~ language.
Generative AI tools were used during development to assist with the creation of certain pre-generated visual and 3D art assets used in the game. AI-assisted content is reviewed, selected, edited, and integrated by the developer. Hollowfall does not generate AI content during gameplay.
The original 1998 version of Avalon was created entirely by hand, including its graphics, music, story, gameplay, game engine and even its custom music software. Avalon: Legacy Edition retains many of these original assets. Generative AI was used to create or refine much of the updated visual, audio and written content. No content is generated live during gameplay.
Card and key art generated with AI tools to a studio style guide, then hand-composited in Photoshop; all text, UI, code and music are original or purchased assets.
All graphics in the game are hand-pixelled. The soundtrack was created with generative AI. Some images for the Steam store and library were expanded using AI to fit the required aspect ratios. No content is generated by AI during gameplay.
I am a solo developer with no budget. This game (my first) has been a passion project of about two years. I have developed it in between work and life. During that time, I made the decision to use a text-to-speech program to generate some NPC dialogue. Someday I hope to be working with a budget; but for now, since I don't have one, I believe that my current solution delivers the best experience for the player.
Generative AI was used only to create artwork for the Steam store page and marketing materials. No generative AI is used in the game or during gameplay.
Pre-generated AI was used to create background music with Suno, localized in-game text for supported languages, and artwork for the Steam store page, library assets, and other marketing materials. No AI-generated images are used during gameplay. All AI-generated content is created before release and does not change during play. The game does not use live-generated AI.
Some of this game's art was stylized or created with AI image generation tools from original prompts and project reference material: character portraits, tables, backgrounds, and Steam store and community artwork. Initial translations were AI-assisted. All of it was edited and reviewed before shipping. The soundtrack is chip-synth music generated at build time by the game's own code, no audio model and no sampled audio in the shipped files. Several tracks are AI-stylized arrangements of original songs written by a member of the development team; others were composed with AI assistance to our direction. Nothing is generated by AI while you play.
I used an AI coding assistant while building the game, and it helped come up with the achievement names and descriptions, which I edited. Nothing else in the game is AI generated: the soundtrack is licensed from its creators, the radio voices are my own recordings, and the art and sounds are produced by the game's own code.
Studs Up! ships some pre-generated AI content. All of it was reviewed and edited by hand before it went into the game:
- Player portraits and the end-of-season win/loss backdrops: image generation, converted down to the game's limited-palette pixel formats.
- The match music loop and the match sound effects: audio generation, edited and loudness-normalised.
- The announcer voice in the trailer: voice generation.
No AI runs while you play. The game is fully offline and generates nothing at runtime.
The current soundtrack and its album-cover artwork were generated with AI music and image tools and selected by the developer. I am looking to commission original music from human artists; if you make music and would like to hear your tracks in BrickAmp, get in touch at team@brickamp.com. Some level layouts and visual effects were drafted with AI-assisted coding tools and hand-reviewed. All of this content is pre-generated and ships with the game. Nothing is generated live during play, and the game does not connect to any AI service.
The game's code, which procedurally generates all 3D models, textures, sound effects and music at runtime, was written with the help of an AI coding assistant, as were the store description, marketing images and trailer. No AI generation happens while the game is running.
AI was used to assist with coding and to create some textures, which were later reworked by the developer.
The game does not generate AI content in real time
AI tools were used to assist in creating selected promotional store assets and game assets. These materials were or will be later edited, refined, and finalized by human creators.
HELIOBANE was made with the help of generative AI tools.
- Assets: AI was used to generate some of the game's assets, including artwork, voice lines, sound effects, in-game text and translations. Artwork was edited and fitted to the game's 32-colour palette and style before use.
- Music: the soundtrack was made with a custom tracker built for agentic work, by an AI trained on a curated collection of public-domain tracker (MOD) music.
- Some store page and marketing art is also AI-assisted.
Nothing is generated while you play. All content is fixed in the build, and the game does not connect to any AI service.
This game uses a mix of human and generative AI tools. AI was used in some visual art assets, code generation, and portions of the store page text. All writing in the game, voice acting, and music are fully human-created.
Dicebound uses generative AI as part of the development process for some pre-generated content. This includes certain visual assets and promotional/store materials, as well as some drafting support for interface text and localization. All AI-assisted content has been reviewed, edited, and curated by the developer before being included in the game or on the store page. Dicebound does not use live generative AI during gameplay, and players cannot generate content through AI in-game.
Gród is designed, built and tested by a solo developer. Every rule, system and pixel of the game follows my own vision.
AI tools helped with parts of the production: the voice lines, the music and some store artwork were created with generative AI, then selected, edited and fitted into the game by me. The in-game pixel art is drawn by the game's own code, with no AI image generators.
Screenshots and GIFs show real gameplay. The game does not use generative AI while you play.
This game uses pre-generated AI-assisted visual content. Generative AI tools were used during development to help create and refine some 2D artwork and Steam store materials, including Chinese ink-style backgrounds, capsule images, UI/marketing artwork, icons, and related visual elements. All AI-assisted assets were reviewed and edited by the developer before being included. The game does not use live generative AI during gameplay, and no player input is sent to an AI service to generate content.
The game engine is written with substantial AI assistance.
Some dialogues and quest texts are AI generated, and are successively replaced by human-written dialogue. Every piece of flavor text has a label that states the author (human or ai).
The 3D character models are currently generated with AI, and then transformed through a custom pipeline to work with the text-based graphics and CPU rendering. It is possible to replace them with manually crafted models.
Some background environment props and their textures (signage, lamps,
surfaces, decals and small set-dressing objects) were generated with AI
assistance, using AI written procedural code rather than an image or mesh
generator. They ship with the game as scenery.
Everything else is the developer's own work: all writing and in-game text,
game design, game code, audio and sound design, the visual identity, and all
store page art.
No AI generated content is produced while the game is running.
DEADLUCK V was created using generative AI, including its artwork, music, sound effects and written content. This content was generated during development. The game does not generate AI content or connect to AI services during gameplay.
This game includes assets created with generative AI during development (official music, song artwork, and drum animations). Nothing is generated while you play, and the game does not connect to an AI service at runtime. Players cannot send prompts to any AI.
Some store and library capsule art was produced with image-generation tools
(SDXL and a chat assistant) using our own in-game screenshots as the input
image. The underlying game content -- models, textures, code, text -- is
authored by us without generative AI. All generated art was reviewed by us
before publication.
AI models have been used to validate, optimize, rank and test the puzzles and the simulation engine. Soundtrack is procedurally generated music using the same engine that simulates the electronics, powered by the C~ language.
Generative AI tools were used during development to assist with the creation of certain pre-generated visual and 3D art assets used in the game. AI-assisted content is reviewed, selected, edited, and integrated by the developer. Hollowfall does not generate AI content during gameplay.
The original 1998 version of Avalon was created entirely by hand, including its graphics, music, story, gameplay, game engine and even its custom music software. Avalon: Legacy Edition retains many of these original assets. Generative AI was used to create or refine much of the updated visual, audio and written content. No content is generated live during gameplay.
Card and key art generated with AI tools to a studio style guide, then hand-composited in Photoshop; all text, UI, code and music are original or purchased assets.
All graphics in the game are hand-pixelled. The soundtrack was created with generative AI. Some images for the Steam store and library were expanded using AI to fit the required aspect ratios. No content is generated by AI during gameplay.
I am a solo developer with no budget. This game (my first) has been a passion project of about two years. I have developed it in between work and life. During that time, I made the decision to use a text-to-speech program to generate some NPC dialogue. Someday I hope to be working with a budget; but for now, since I don't have one, I believe that my current solution delivers the best experience for the player.
Generative AI was used only to create artwork for the Steam store page and marketing materials. No generative AI is used in the game or during gameplay.
Pre-generated AI was used to create background music with Suno, localized in-game text for supported languages, and artwork for the Steam store page, library assets, and other marketing materials. No AI-generated images are used during gameplay. All AI-generated content is created before release and does not change during play. The game does not use live-generated AI.
Some of this game's art was stylized or created with AI image generation tools from original prompts and project reference material: character portraits, tables, backgrounds, and Steam store and community artwork. Initial translations were AI-assisted. All of it was edited and reviewed before shipping. The soundtrack is chip-synth music generated at build time by the game's own code, no audio model and no sampled audio in the shipped files. Several tracks are AI-stylized arrangements of original songs written by a member of the development team; others were composed with AI assistance to our direction. Nothing is generated by AI while you play.
I used an AI coding assistant while building the game, and it helped come up with the achievement names and descriptions, which I edited. Nothing else in the game is AI generated: the soundtrack is licensed from its creators, the radio voices are my own recordings, and the art and sounds are produced by the game's own code.
Studs Up! ships some pre-generated AI content. All of it was reviewed and edited by hand before it went into the game:
- Player portraits and the end-of-season win/loss backdrops: image generation, converted down to the game's limited-palette pixel formats.
- The match music loop and the match sound effects: audio generation, edited and loudness-normalised.
- The announcer voice in the trailer: voice generation.
No AI runs while you play. The game is fully offline and generates nothing at runtime.
