Public Steam disclosure research26,832 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.
During development, we used some AI tools to speed things up in areas where we weren’t as experienced. That saved us a lot of time, letting us focus on the fun stuff — creating cool gameplay mechanics, improving performance, and making sure the game runs smoothly without annoying bugs.
1.The game utilizes AI-assisted ideas for programming logic to accelerate the development process.
2.Current music was created with the help of an AI-based platform.
3. This game does not use any form of AI for image generation.
we use AI across multiple areas of our creative process. This includes generating and enhancing textures, videos, stories, logos, concepts and ideas, music, voice acting, sound effects, and even 3D models. AI tools help us experiment faster, explore new artistic directions, and bring our worlds to life with greater efficiency and creativity.
All visual assets within the application, including icons, backgrounds, and UI elements, were created using generative AI tools during the development process. These assets are pre-rendered and static. The application does not utilize AI to generate content in real-time.
The mission dialogues and character voices were created using AI. Since the game is fully offline, the voices are pre-generated and not produced in real time during gameplay.
Umbra Survivors does not use artificial intelligence during runtime.
AI tools were only used as development assistance to help prepare, optimize, or comment on code.
All in-game content (graphics, sounds, text, and code) has been manually reviewed and originates from internal development or properly licensed sources.
Some image assets are generated or edited using these tools, but all illustrations are ultimately created by an artist. Furthermore, these tools were used to produce animations for the trailer.
This game uses generative AI tools to create pre-rendered content during development, specifically applied in the following aspects:
Art assets: Scene concept art, character design sketches, and basic UI icon materials. All AI-generated visual content has undergone manual review, modification, and optimization to ensure consistent style and no infringement risks.
Marketing and community assets: Some background images on the store page, auxiliary elements of promotional posters, and non-core visual materials for community sharing, all complying with Steam's content guidelines.
Development assistance: Prototypes of certain sound effects and initial drafts of code snippets, all of which have passed manual verification and adjustment to ensure functional compliance and stability.
All AI-generated content has undergone strict intellectual property verification to ensure no illegal or infringing information is included, and the actual game presentation is consistent with marketing materials.
Use of AI in Development (Pre-Generated Only)
-Our game does not create new content at runtime. During development we used AI as a helper in two ways:
-Programming: code-completion and refactor suggestions (e.g., test scaffolds, lint fixes). All code was reviewed and modified by a human before inclusion.
-Assets: early concepts/placeholders for UI icons and marketing drafts. Final in-game and store assets were painted/edited and approved by our team before shipping.
-We do not ship unreviewed model outputs, and we do not include content we believe infringes others’ rights. This disclosure follows Steam’s guidance to label pre-generated vs. live-generated AI usage; our title uses pre-generated only.
I am a solo developer building a massive game. This naturally presents major time and budget challenges for creating all the art assets.
To maintain my ambitious visual scope, I utilize AI-generated concepts purely as a starting foundation. I personally hand-paint and finalize every draft with digital painting before it enters the game. This hybrid art procedure granted me the speed I needed also helped maintaining the visual artistic standards.
The app was developed through both my own coding effort, as well as back and forth coding prompts with ChatGPT 5. Additionally, art assets were generated through use of ChatGPT 5.
Only AI Code was used.
- Used to create C# scripts for the following features: floating object direction calculations, boss camera, replacing items in shop interface, porting over attack hitbox logic from laggy events into C# code, lip-sync implementation, online leaderboard send/receive, dynamic button swap with keybind update, proximity interaction UI call implementation, mouse/controller sensitivity options, dropped weapon properties.
- Used to create external tools to speed up development: blender plug-ins, web tools to parse and modify event data, python scripts.
- Used for debugging purposes.
During the development of this game, generative AI was used solely to assist in the creation of certain 3D models. All core gameplay design, and narrative content were created independently by our development team.
This game does not use generative AI in real time. Generative models were used only during early development to create draft materials (such as texture sketches, concept art, or music prototypes), which were then manually refined. All final assets are original and fully adapted for the game.
There's one painting inside the mansion building, in chapter 2 called "Full House" chapter.
That painting was generated by OpenAI ChatGpt as texture for painting mesh.
it's not live-generate content, but it's pre-rendered by AI.
Players can optionally use AI-assisted world generation when creating new game worlds. Through a structured wizard
interface, players configure:
- World name and theme (e.g., "Classic Fantasy")
- Geographic bounds
- Optional location suggestions
The AI then generates points of interest (shrines, inns, guild halls, ruins), NPC characters with names and
personalities, dialogue (greetings, backstories, quest hooks), and story assignments.
Players do not have direct chat access to the AI. All generation flows through fixed game parameters and moderated
output. The feature is optional - players can also create worlds manually.
AI is not used for any major assets or parts of Ludicrum's development. The limited usage was for a few custom generated, niche textures that could not be easily produced.
Some in-game text were proofread and improved by AI-powered writing reviewer.
No other game assets were AI-gnereated. All sprites you see here are hand-pixelated.
During development, we used some AI tools to speed things up in areas where we weren’t as experienced. That saved us a lot of time, letting us focus on the fun stuff — creating cool gameplay mechanics, improving performance, and making sure the game runs smoothly without annoying bugs.
1.The game utilizes AI-assisted ideas for programming logic to accelerate the development process.
2.Current music was created with the help of an AI-based platform.
3. This game does not use any form of AI for image generation.
we use AI across multiple areas of our creative process. This includes generating and enhancing textures, videos, stories, logos, concepts and ideas, music, voice acting, sound effects, and even 3D models. AI tools help us experiment faster, explore new artistic directions, and bring our worlds to life with greater efficiency and creativity.
All visual assets within the application, including icons, backgrounds, and UI elements, were created using generative AI tools during the development process. These assets are pre-rendered and static. The application does not utilize AI to generate content in real-time.
The mission dialogues and character voices were created using AI. Since the game is fully offline, the voices are pre-generated and not produced in real time during gameplay.
Umbra Survivors does not use artificial intelligence during runtime.
AI tools were only used as development assistance to help prepare, optimize, or comment on code.
All in-game content (graphics, sounds, text, and code) has been manually reviewed and originates from internal development or properly licensed sources.
Some image assets are generated or edited using these tools, but all illustrations are ultimately created by an artist. Furthermore, these tools were used to produce animations for the trailer.
This game uses generative AI tools to create pre-rendered content during development, specifically applied in the following aspects:
Art assets: Scene concept art, character design sketches, and basic UI icon materials. All AI-generated visual content has undergone manual review, modification, and optimization to ensure consistent style and no infringement risks.
Marketing and community assets: Some background images on the store page, auxiliary elements of promotional posters, and non-core visual materials for community sharing, all complying with Steam's content guidelines.
Development assistance: Prototypes of certain sound effects and initial drafts of code snippets, all of which have passed manual verification and adjustment to ensure functional compliance and stability.
All AI-generated content has undergone strict intellectual property verification to ensure no illegal or infringing information is included, and the actual game presentation is consistent with marketing materials.
Use of AI in Development (Pre-Generated Only)
-Our game does not create new content at runtime. During development we used AI as a helper in two ways:
-Programming: code-completion and refactor suggestions (e.g., test scaffolds, lint fixes). All code was reviewed and modified by a human before inclusion.
-Assets: early concepts/placeholders for UI icons and marketing drafts. Final in-game and store assets were painted/edited and approved by our team before shipping.
-We do not ship unreviewed model outputs, and we do not include content we believe infringes others’ rights. This disclosure follows Steam’s guidance to label pre-generated vs. live-generated AI usage; our title uses pre-generated only.
I am a solo developer building a massive game. This naturally presents major time and budget challenges for creating all the art assets.
To maintain my ambitious visual scope, I utilize AI-generated concepts purely as a starting foundation. I personally hand-paint and finalize every draft with digital painting before it enters the game. This hybrid art procedure granted me the speed I needed also helped maintaining the visual artistic standards.
The app was developed through both my own coding effort, as well as back and forth coding prompts with ChatGPT 5. Additionally, art assets were generated through use of ChatGPT 5.
Only AI Code was used.
- Used to create C# scripts for the following features: floating object direction calculations, boss camera, replacing items in shop interface, porting over attack hitbox logic from laggy events into C# code, lip-sync implementation, online leaderboard send/receive, dynamic button swap with keybind update, proximity interaction UI call implementation, mouse/controller sensitivity options, dropped weapon properties.
- Used to create external tools to speed up development: blender plug-ins, web tools to parse and modify event data, python scripts.
- Used for debugging purposes.
During the development of this game, generative AI was used solely to assist in the creation of certain 3D models. All core gameplay design, and narrative content were created independently by our development team.
This game does not use generative AI in real time. Generative models were used only during early development to create draft materials (such as texture sketches, concept art, or music prototypes), which were then manually refined. All final assets are original and fully adapted for the game.
There's one painting inside the mansion building, in chapter 2 called "Full House" chapter.
That painting was generated by OpenAI ChatGpt as texture for painting mesh.
it's not live-generate content, but it's pre-rendered by AI.
Players can optionally use AI-assisted world generation when creating new game worlds. Through a structured wizard
interface, players configure:
- World name and theme (e.g., "Classic Fantasy")
- Geographic bounds
- Optional location suggestions
The AI then generates points of interest (shrines, inns, guild halls, ruins), NPC characters with names and
personalities, dialogue (greetings, backstories, quest hooks), and story assignments.
Players do not have direct chat access to the AI. All generation flows through fixed game parameters and moderated
output. The feature is optional - players can also create worlds manually.
AI is not used for any major assets or parts of Ludicrum's development. The limited usage was for a few custom generated, niche textures that could not be easily produced.
Some in-game text were proofread and improved by AI-powered writing reviewer.
No other game assets were AI-gnereated. All sprites you see here are hand-pixelated.