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.
Stay With Me contains a limited amount of pre-generated content created with the assistance of generative AI.
Some visual assets, the loading screen artwork, and some voice lines for the characters Margaret and the child were created using AI tools.
The AI-generated content was further edited, processed and adapted by the developer before being integrated into the game.
All AI content is pre-generated during development. The game does not generate content using AI while running, and there is no live AI generation.
We use generative AI tools during development to create some concept art and supporting images, as well as certain materials for the store page. All in-game content, including final assets, models, and text, is created and reviewed by hand.
Some artwork for the game or its store materials includes minor AI-assisted edits. This content is pre-generated. The game does not generate AI content during gameplay and does not connect to external AI services.
An AI assistant (Claude) was used during development. It wrote the in-game text translations, the store page text, and the scripts that build many of the game's 3D models, animations and textures, all under the developer's direction and review. Nothing is generated by AI while you play. All sounds are real recordings from CC0 libraries.
A small language model (Google's Gemma 3 270M) runs entirely on the player's computer, with no network connection. It writes short lines of villager small talk, the ritual oracle's one- or two-sentence hint after a failed ritual, and routine filler records in the in-game archive, and it chooses which of our pre-written dialogue lines fits the moment.
Guardrails: the player never types to the model — there is no text input anywhere in the game. Every prompt is a fixed template we wrote, filled in from game state such as a character's name, the time of day and the current quest. Output is short by construction (hard token caps, and lines over a few words are cut or discarded); lines that read as narration, stage directions or an assistant speaking are rejected; dialogue choices are constrained to our own authored lines. Any failure falls back to written lines. Players can turn the model off in Settings, and the game plays identically on written lines alone.
Generative AI was used during development to create select environmental artwork, including some fictional posters and signage appearing within the game. These assets were reviewed and curated by the developers before inclusion in the final game.
Part of the game's text, including the German, French and Spanish translations, and the store page descriptions were written with the help of AI tools. The paintings are public-domain works by real artists.
Generative AI was used to create the illustrations, as well as a small portion of the character artwork and dialogue. This content was created during development. The game does not generate AI content during gameplay.
Brine & Brass is a solo project by Remoria, assisted by AI. GPT assisted with game programming, in-game writing, artwork-drawing scripts, and procedural music and sound code. The artwork is rendered from project drawing code, and the audio uses local waveform synthesis. No generative AI model or online AI service runs during gameplay.
This game utilizes generative AI for a significant portion of its art assets. Many AI-assisted elements underwent manual editing and modification by the developer before being integrated into the final game.
Some 3D models in this game, with their textures, were generated with an AI 3D-modeling tool from our own designs and then rigged and edited for the game: the Harvester, the Halloway family, the two playable co-op characters, the scarecrows and the wax effigy.
We use generative AI across all disciplines of game development, most often with heavy amounts of human editing. That involves audio and visual art, 3D models, our logo, and localization. We use it as a productivity tool so we can achieve a higher level of quality, faster, than we would have been able to.
Generative AI was used to help create some of the promotional artwork and marketing images displayed on Station 12’s Steam store page.
These images are promotional illustrations and are a good representation of the in-game visuals.
AI is used to assist with the localization and translation of the game's store page and in game text. It is also used to assist with some simple voice interactions, such as goblin growls and battle cries, as well as basic texture adjustments that fit the game's low poly visual style.
This game was made with the help of generative AI.
AI tools were used during development to create visual assets, sound and music, parts
of the source code, and the translations of this page and the in-game text. Everything
was chosen, edited and assembled by hand before release.
Nothing is generated while you play — all content ships with the game.
Some of this game's art, music, sound effects, 3D models, and translations were created with the help of generative AI. No content is generated by AI during gameplay.
Our visual style relies on a blend of commissioned artwork, curated assets, and custom designs. We occasionally use AI generation as a tool in our creative process, but never as a final product. Every AI-assisted asset is either fully redrawn or heavily refined by human artists before making its way into the game.
Store page and marketing: the main capsule art was composed and rendered by the developer from in-game models, with generative AI used to repaint parts of the background (landscape, sky and lighting), then corrected by hand. Some promotional videos use an AI-generated voiceover.
Misty and the Last Breeze is a solo-developed game made with a lot of AI assistance. Here is how AI was used.
Code. The game is built in Godot. Most of the gameplay code, tooling and automated tests were written with AI coding assistants (Claude Code and Codex), working from my design direction, and reviewed, playtested and revised by me. Design decisions, level layouts, movement feel and what ships are mine; the assistants do much of the typing.
3D models and art. Many in-game models and textures were made in Blender through scripts written with AI assistance, often rebuilt from concept images. Some of those reference images were generated with AI image tools. Other visuals, such as shaders, UI and effects, are authored in code. Painting and key-art images in the game that began as AI-generated images are kept as AI-assisted work.
Music. Music is generated with Suno and then selected, arranged and edited by me. Ambience and sound effects are mostly synthesized in code (procedural audio) rather than sampled from other artists.
Mockups and planning. I use ChatGPT's image generation for mockups, mood boards, UI layouts and promotional concepts while planning. These are for exploring ideas; they are not meant to ship as finished art unless noted above.
What is not AI. The game itself has no runtime AI: every enemy, NPC, line of dialogue and level is fixed content that I designed and authored. Nothing is generated or adapted while you play.
Generative AI was used to create some art assets (the title screen background and the store capsule images). The game does not use AI to generate any content at runtime.
Generative AI-assisted tools were used during development to support the creation and refinement of some pre-generated visual assets. No AI content is generated during gameplay.
Some 3D models, illustrations and music in the current version were created with the help of generative AI tools and then edited and integrated by our team. They are placeholders: we are replacing them with work by human artists as development continues.
The key art, store capsules and library artwork were created with an AI image generator and edited by the developer. The trailer music was generated with Suno under a paid plan. No AI-generated content is included in the game itself, and the game runs no AI generation and requires no AI service.
Generative AI tools were used during development to create visual assets, including character and enemy artwork, environmental backgrounds, user interface graphics, item textures, and promotional artwork. ChatGPT also assisted with dialogue and educational text. The developer directed, selected, and reviewed this content for inclusion in the game. All AI-generated content was created before release; the game does not generate AI content during gameplay.
Generative AI was used during development to create and refine some of the game's 2d and 3d artwork and music. All AI-generated content was reviewed, selected, and edited by the developer before being included in the game. Generative AI is not used during gameplay, and no content is generated in real time.
Some of Grim Island's artwork was created using generative AI image tools, working from original artwork hand-drawn by the developer and then curated and edited by the developer. No other content is AI-generated.
We used AI-assisted tools to help create and refine a small selection of visual assets. Each one was reviewed and adjusted by hand before it was used.
AI also helped us tidy up the grammar and wording in a small amount of in-game and store-page text. We checked and edited every final line ourselves.
For us, AI is simply a supporting tool. Dino Dome: Tidy Up! is still shaped by our own ideas, judgement, and hands-on creative work.
AI tools were used to create or edit some in-game and store artwork and textures, and to assist with translations. Creative direction and final content selection are ours. No AI content is generated during gameplay.
Stay With Me contains a limited amount of pre-generated content created with the assistance of generative AI.
Some visual assets, the loading screen artwork, and some voice lines for the characters Margaret and the child were created using AI tools.
The AI-generated content was further edited, processed and adapted by the developer before being integrated into the game.
All AI content is pre-generated during development. The game does not generate content using AI while running, and there is no live AI generation.
We use generative AI tools during development to create some concept art and supporting images, as well as certain materials for the store page. All in-game content, including final assets, models, and text, is created and reviewed by hand.
Some artwork for the game or its store materials includes minor AI-assisted edits. This content is pre-generated. The game does not generate AI content during gameplay and does not connect to external AI services.
An AI assistant (Claude) was used during development. It wrote the in-game text translations, the store page text, and the scripts that build many of the game's 3D models, animations and textures, all under the developer's direction and review. Nothing is generated by AI while you play. All sounds are real recordings from CC0 libraries.
A small language model (Google's Gemma 3 270M) runs entirely on the player's computer, with no network connection. It writes short lines of villager small talk, the ritual oracle's one- or two-sentence hint after a failed ritual, and routine filler records in the in-game archive, and it chooses which of our pre-written dialogue lines fits the moment.
Guardrails: the player never types to the model — there is no text input anywhere in the game. Every prompt is a fixed template we wrote, filled in from game state such as a character's name, the time of day and the current quest. Output is short by construction (hard token caps, and lines over a few words are cut or discarded); lines that read as narration, stage directions or an assistant speaking are rejected; dialogue choices are constrained to our own authored lines. Any failure falls back to written lines. Players can turn the model off in Settings, and the game plays identically on written lines alone.
Generative AI was used during development to create select environmental artwork, including some fictional posters and signage appearing within the game. These assets were reviewed and curated by the developers before inclusion in the final game.
Part of the game's text, including the German, French and Spanish translations, and the store page descriptions were written with the help of AI tools. The paintings are public-domain works by real artists.
Generative AI was used to create the illustrations, as well as a small portion of the character artwork and dialogue. This content was created during development. The game does not generate AI content during gameplay.
Brine & Brass is a solo project by Remoria, assisted by AI. GPT assisted with game programming, in-game writing, artwork-drawing scripts, and procedural music and sound code. The artwork is rendered from project drawing code, and the audio uses local waveform synthesis. No generative AI model or online AI service runs during gameplay.
This game utilizes generative AI for a significant portion of its art assets. Many AI-assisted elements underwent manual editing and modification by the developer before being integrated into the final game.
Some 3D models in this game, with their textures, were generated with an AI 3D-modeling tool from our own designs and then rigged and edited for the game: the Harvester, the Halloway family, the two playable co-op characters, the scarecrows and the wax effigy.
We use generative AI across all disciplines of game development, most often with heavy amounts of human editing. That involves audio and visual art, 3D models, our logo, and localization. We use it as a productivity tool so we can achieve a higher level of quality, faster, than we would have been able to.
Generative AI was used to help create some of the promotional artwork and marketing images displayed on Station 12’s Steam store page.
These images are promotional illustrations and are a good representation of the in-game visuals.
AI is used to assist with the localization and translation of the game's store page and in game text. It is also used to assist with some simple voice interactions, such as goblin growls and battle cries, as well as basic texture adjustments that fit the game's low poly visual style.
This game was made with the help of generative AI.
AI tools were used during development to create visual assets, sound and music, parts
of the source code, and the translations of this page and the in-game text. Everything
was chosen, edited and assembled by hand before release.
Nothing is generated while you play — all content ships with the game.
Some of this game's art, music, sound effects, 3D models, and translations were created with the help of generative AI. No content is generated by AI during gameplay.
Our visual style relies on a blend of commissioned artwork, curated assets, and custom designs. We occasionally use AI generation as a tool in our creative process, but never as a final product. Every AI-assisted asset is either fully redrawn or heavily refined by human artists before making its way into the game.
Store page and marketing: the main capsule art was composed and rendered by the developer from in-game models, with generative AI used to repaint parts of the background (landscape, sky and lighting), then corrected by hand. Some promotional videos use an AI-generated voiceover.
Misty and the Last Breeze is a solo-developed game made with a lot of AI assistance. Here is how AI was used.
Code. The game is built in Godot. Most of the gameplay code, tooling and automated tests were written with AI coding assistants (Claude Code and Codex), working from my design direction, and reviewed, playtested and revised by me. Design decisions, level layouts, movement feel and what ships are mine; the assistants do much of the typing.
3D models and art. Many in-game models and textures were made in Blender through scripts written with AI assistance, often rebuilt from concept images. Some of those reference images were generated with AI image tools. Other visuals, such as shaders, UI and effects, are authored in code. Painting and key-art images in the game that began as AI-generated images are kept as AI-assisted work.
Music. Music is generated with Suno and then selected, arranged and edited by me. Ambience and sound effects are mostly synthesized in code (procedural audio) rather than sampled from other artists.
Mockups and planning. I use ChatGPT's image generation for mockups, mood boards, UI layouts and promotional concepts while planning. These are for exploring ideas; they are not meant to ship as finished art unless noted above.
What is not AI. The game itself has no runtime AI: every enemy, NPC, line of dialogue and level is fixed content that I designed and authored. Nothing is generated or adapted while you play.
Generative AI was used to create some art assets (the title screen background and the store capsule images). The game does not use AI to generate any content at runtime.
Generative AI-assisted tools were used during development to support the creation and refinement of some pre-generated visual assets. No AI content is generated during gameplay.
Some 3D models, illustrations and music in the current version were created with the help of generative AI tools and then edited and integrated by our team. They are placeholders: we are replacing them with work by human artists as development continues.
The key art, store capsules and library artwork were created with an AI image generator and edited by the developer. The trailer music was generated with Suno under a paid plan. No AI-generated content is included in the game itself, and the game runs no AI generation and requires no AI service.
Generative AI tools were used during development to create visual assets, including character and enemy artwork, environmental backgrounds, user interface graphics, item textures, and promotional artwork. ChatGPT also assisted with dialogue and educational text. The developer directed, selected, and reviewed this content for inclusion in the game. All AI-generated content was created before release; the game does not generate AI content during gameplay.
Generative AI was used during development to create and refine some of the game's 2d and 3d artwork and music. All AI-generated content was reviewed, selected, and edited by the developer before being included in the game. Generative AI is not used during gameplay, and no content is generated in real time.
Some of Grim Island's artwork was created using generative AI image tools, working from original artwork hand-drawn by the developer and then curated and edited by the developer. No other content is AI-generated.
We used AI-assisted tools to help create and refine a small selection of visual assets. Each one was reviewed and adjusted by hand before it was used.
AI also helped us tidy up the grammar and wording in a small amount of in-game and store-page text. We checked and edited every final line ourselves.
For us, AI is simply a supporting tool. Dino Dome: Tidy Up! is still shaped by our own ideas, judgement, and hands-on creative work.
AI tools were used to create or edit some in-game and store artwork and textures, and to assist with translations. Creative direction and final content selection are ours. No AI content is generated during gameplay.
