Public Steam disclosure research25,383 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.
Many of the game images have been made with AI though they have been heavily edited afterwards. It has also assisted with some of the code. This game would not be possible without it as I'm a small indie dev.
Given I am a solo developer, Generative AI tools were utilized as development assistants to optimize workflows.
Also, some of the backgrounds are based on images created with AI, as well as a few items. They were only used as a basis to start from and have been heavily edited manually afterwards to make them fit with the rest of the game's pixel art artstyle.
I have used AI to design some parts of the main menu, data models, design of the multiplayer system and room system and to improve the heuristic armies behavior.
During development, this game uses the following generative AI tools. All generated assets are reviewed, selected, and adjusted by the developer before being incorporated into the game.
Music (BGM): Composed using Suno
Battle animations: Generated using Seedance 2.0
Image assets (characters, backgrounds, etc.): Generated using GPT Image 2.0
Audio/Voice assets: Irodori-TTS (AI text-to-speech service)
All of these are pre-generated content created during development. No AI-generated content is produced at runtime (live) during gameplay.
All content generated by the above tools is licensed for commercial use in accordance with each service's terms of use, and does not infringe upon any third-party copyrights or other intellectual property rights.
This title makes assistive use of generative AI tools in the following areas. All AI-generated outputs are manually reviewed, edited, and curated by the development team before being incorporated as final assets, and we have taken care to ensure that they do not infringe on the rights of third parties. No real-time AI generation occurs during gameplay.[Pre-Rendered Content — AI-Assisted Areas]- Production support for a portion of the in-game visual assets- Production support for a portion of the in-game music (BGM) and sound effects- Translation support for store page localization across supported languages (with human proofreading)[Live Generation]The game does not invoke generative AI in real time at any point during gameplay.
Generative AI tools were used during development for ideation, development assistance, and promotional exploration. All final gameplay systems, game logic, and in-game content were created and reviewed by the developer.
Master of the Ages is a grand strategy game engineered by a solo developer with 15+ years of senior software engineering and programming competition experience. To achieve grand-strategy scale as a single creator, a generative AI workflow was utilized to co-design the code architecture and write the codebase under the developer's supervision, backed by a rigorous framework of over 2,500 automated tests to catch edge-case bugs and prevent regressions during development. Generative AI was also used to produce all player-facing graphics (including all unit, building, and technology assets) and generate the historical text layers.
Crown Caps Soccer uses generative artificial intelligence tools for part of its content. Specifically, we use AI to generate and translate interface text and descriptions, as well as to assist in creating certain in-game content. All AI-generated content is reviewed by our team before publication.
Ten Rings 2 uses generative AI during development to assist with some 2D visual asset production, concept iteration, and writing drafts. Final in-game content is selected, edited, integrated, and reviewed by the developer. Generative AI is not used to create live or personalized content during gameplay.
During the development of Super Monsterdoom, generative AI tools were used to support text creation, visual exploration, and mechanical iteration. As a solo developer, these tools were key to accelerating and expanding the process, but every final asset, text, and design decision went through extensive human editing, refinement, and hands-on work.
Some marketing artwork on this store page (capsule images, library hero, page background) was generated with AI image tools and then composited with vector-rendered typography. No AI-generated content ships inside the game itself. All player names, team names, draft classes, and simulation outcomes are produced by procedural algorithms with a seeded random number generator, not by AI models.
- Code was written with Assistance of Anthropic Claude Code
- Character Portraits were created with Google Gemini
- Other Art Assets and Sound Effects were created using Python (matplotlib, wave)
Pre-generated AI-assisted artwork is used for fictional chess player avatars and selected promotional visuals. These assets are reviewed before inclusion. No live generative AI is used during gameplay.
Delphi uses Midjourney for illustrated artwork (cards, events, key art, capsule art) and Kling for image-to-video trailer animation, all curated and edited by the developer. UI artwork is hand-made by a credited human artist.
No AI is used in design, writing, narrative or gameplay, and no AI runs at runtime — all in-game content is human-written static data.
We used ai to assist us on music and art because we are poor and not capable of this yet. We will strive to remove or reduce ai to a minimum once we are financially capable.
Some visual assets used in The Cornerman, including certain fighter portraits and background artwork, were created or assisted using generative AI tools and were further edited and integrated by the developer before inclusion in the game.
Some vfx sprite sheets and art were generated and then photoshopped and shaders added. I generated some images 100 plus times to make sure the style didn't drift. I photoshopped AI images to get what I wanted. AI generated images were guided by my thoughts the entire time.
AI tools were used during development only for pre-generated content, including concept/reference artwork, store assets, music/audio drafts, and localization or translation assistance.
The game does not generate AI content at runtime. Players do not interact with any live AI system, and no text, images, audio, or gameplay content is dynamically generated by AI while the game is running.
All AI-assisted content is reviewed, edited, and integrated by the developer before release.
Generative AI tools are used in this project. Claude and ChatGPT assist with drafting store page copy and marketing media. Some in-game visual assets are created with AI image generation.
Some store artwork and promotional images were created or refined with AI-assisted tools, then edited and curated by the developer. The game does not include live AI-generated content.
AI assisted assets and text are developer edited. The core game is playable without AI. Optional advisor commentary can use local Ollama or Anthropic only if the player enables it and provides their own key.
Some visual assets used on this Steam page were edited or retouched using AI-assisted tools, and the achievement icons were created using generative AI / AI-assisted tools.
As a small indie studio, we use generative AI tools to assist in parts of our art creation process. This helps us improve the quality and consistency of artwork while allowing our team to work more efficiently. All AI-assisted content is reviewed, refined, and integrated by our developers to ensure it meets the artistic vision and quality standards of the game.
Some UI icons, character and creature sprites, and other in-game graphics in Submission Saga were created with the assistance of generative AI image tools (including Pixellab.ai), and portions of the store page text were drafted with AI writing assistance. Portions of the game's source code were also written with the assistance of an AI coding assistant. All AI-generated content was reviewed, edited, and curated by the developer before inclusion. No AI is used to generate content live during gameplay.
Translations and text content are AI-generated. Some of the UI and game backgrounds are AI generated with manual post-processing. Some of the VFX for the animations are AI generated with post-processing.
Some of the text descriptions on the game’s store page, certain in-game localization texts, and some images used in the test version were generated with the help of AI technology.
The static pixel art images of the creatures (Deeplords) and environmental elements were pre-generated using AI tools and subsequently animated with another AI tool based on the developer's specific directions. Leveraging 10 years of professional 3D modeling experience, these assets were then integrated as textures and sprites inside complex 3D environments and dioramas modeled manually by the developer in Blender and structured in Unity. The game's Unity C# scripts were pre-generated using AI co-pilot tools based on the precise gameplay logic requested by the developer, followed by extensive manual debugging, correction, and multi-turn optimization cycles. Lastly, the soundtrack and audio elements were pre-generated using AI tools based on specific genre and stylistic vibes provided by the developer, then manually reviewed and refined to perfectly fit the dark atmosphere of the individual biomes.
This game uses generative artificial intelligence (AI) to create some of its visual art assets. Certain in-game images and illustrations were partially generated with the assistance of AI-based painting tools.
Kingdom of Ashes uses generative AI for select visual assets:
• Character portraits, enemy sprites, card illustrations, and background art were generated using Microsoft Copilot and then curated, composited, and refined by the developer for use in the game.
All gameplay code, mechanics, narrative text, and game design were created by the developer. No AI content is generated live during gameplay, and the game does not connect to any third-party AI service at runtime — all AI-generated visuals were finalized during development and ship as static assets.
Based on the developer's original creative direction, AI tools were used to generate the game's store page capsule images, soundtrack, and the non-English in-game text translations. AI coding assistance supported programming during development.
We're a small indie team, and we used generative AI to help make some of the art in Last Ember — card illustrations, backgrounds, character portraits, and a few short animations. We also used it on a handful of sound effects. Everything you see and hear in the game was picked, edited, and signed off by us. The game itself doesn't run any AI while you play — what ships is what you get.
Many of the game images have been made with AI though they have been heavily edited afterwards. It has also assisted with some of the code. This game would not be possible without it as I'm a small indie dev.
Given I am a solo developer, Generative AI tools were utilized as development assistants to optimize workflows.
Also, some of the backgrounds are based on images created with AI, as well as a few items. They were only used as a basis to start from and have been heavily edited manually afterwards to make them fit with the rest of the game's pixel art artstyle.
I have used AI to design some parts of the main menu, data models, design of the multiplayer system and room system and to improve the heuristic armies behavior.
During development, this game uses the following generative AI tools. All generated assets are reviewed, selected, and adjusted by the developer before being incorporated into the game.
Music (BGM): Composed using Suno
Battle animations: Generated using Seedance 2.0
Image assets (characters, backgrounds, etc.): Generated using GPT Image 2.0
Audio/Voice assets: Irodori-TTS (AI text-to-speech service)
All of these are pre-generated content created during development. No AI-generated content is produced at runtime (live) during gameplay.
All content generated by the above tools is licensed for commercial use in accordance with each service's terms of use, and does not infringe upon any third-party copyrights or other intellectual property rights.
This title makes assistive use of generative AI tools in the following areas. All AI-generated outputs are manually reviewed, edited, and curated by the development team before being incorporated as final assets, and we have taken care to ensure that they do not infringe on the rights of third parties. No real-time AI generation occurs during gameplay.[Pre-Rendered Content — AI-Assisted Areas]- Production support for a portion of the in-game visual assets- Production support for a portion of the in-game music (BGM) and sound effects- Translation support for store page localization across supported languages (with human proofreading)[Live Generation]The game does not invoke generative AI in real time at any point during gameplay.
Generative AI tools were used during development for ideation, development assistance, and promotional exploration. All final gameplay systems, game logic, and in-game content were created and reviewed by the developer.
Master of the Ages is a grand strategy game engineered by a solo developer with 15+ years of senior software engineering and programming competition experience. To achieve grand-strategy scale as a single creator, a generative AI workflow was utilized to co-design the code architecture and write the codebase under the developer's supervision, backed by a rigorous framework of over 2,500 automated tests to catch edge-case bugs and prevent regressions during development. Generative AI was also used to produce all player-facing graphics (including all unit, building, and technology assets) and generate the historical text layers.
Crown Caps Soccer uses generative artificial intelligence tools for part of its content. Specifically, we use AI to generate and translate interface text and descriptions, as well as to assist in creating certain in-game content. All AI-generated content is reviewed by our team before publication.
Ten Rings 2 uses generative AI during development to assist with some 2D visual asset production, concept iteration, and writing drafts. Final in-game content is selected, edited, integrated, and reviewed by the developer. Generative AI is not used to create live or personalized content during gameplay.
During the development of Super Monsterdoom, generative AI tools were used to support text creation, visual exploration, and mechanical iteration. As a solo developer, these tools were key to accelerating and expanding the process, but every final asset, text, and design decision went through extensive human editing, refinement, and hands-on work.
Some marketing artwork on this store page (capsule images, library hero, page background) was generated with AI image tools and then composited with vector-rendered typography. No AI-generated content ships inside the game itself. All player names, team names, draft classes, and simulation outcomes are produced by procedural algorithms with a seeded random number generator, not by AI models.
- Code was written with Assistance of Anthropic Claude Code
- Character Portraits were created with Google Gemini
- Other Art Assets and Sound Effects were created using Python (matplotlib, wave)
Pre-generated AI-assisted artwork is used for fictional chess player avatars and selected promotional visuals. These assets are reviewed before inclusion. No live generative AI is used during gameplay.
Delphi uses Midjourney for illustrated artwork (cards, events, key art, capsule art) and Kling for image-to-video trailer animation, all curated and edited by the developer. UI artwork is hand-made by a credited human artist.
No AI is used in design, writing, narrative or gameplay, and no AI runs at runtime — all in-game content is human-written static data.
We used ai to assist us on music and art because we are poor and not capable of this yet. We will strive to remove or reduce ai to a minimum once we are financially capable.
Some visual assets used in The Cornerman, including certain fighter portraits and background artwork, were created or assisted using generative AI tools and were further edited and integrated by the developer before inclusion in the game.
Some vfx sprite sheets and art were generated and then photoshopped and shaders added. I generated some images 100 plus times to make sure the style didn't drift. I photoshopped AI images to get what I wanted. AI generated images were guided by my thoughts the entire time.
AI tools were used during development only for pre-generated content, including concept/reference artwork, store assets, music/audio drafts, and localization or translation assistance.
The game does not generate AI content at runtime. Players do not interact with any live AI system, and no text, images, audio, or gameplay content is dynamically generated by AI while the game is running.
All AI-assisted content is reviewed, edited, and integrated by the developer before release.
Generative AI tools are used in this project. Claude and ChatGPT assist with drafting store page copy and marketing media. Some in-game visual assets are created with AI image generation.
Some store artwork and promotional images were created or refined with AI-assisted tools, then edited and curated by the developer. The game does not include live AI-generated content.
AI assisted assets and text are developer edited. The core game is playable without AI. Optional advisor commentary can use local Ollama or Anthropic only if the player enables it and provides their own key.
Some visual assets used on this Steam page were edited or retouched using AI-assisted tools, and the achievement icons were created using generative AI / AI-assisted tools.
As a small indie studio, we use generative AI tools to assist in parts of our art creation process. This helps us improve the quality and consistency of artwork while allowing our team to work more efficiently. All AI-assisted content is reviewed, refined, and integrated by our developers to ensure it meets the artistic vision and quality standards of the game.
Some UI icons, character and creature sprites, and other in-game graphics in Submission Saga were created with the assistance of generative AI image tools (including Pixellab.ai), and portions of the store page text were drafted with AI writing assistance. Portions of the game's source code were also written with the assistance of an AI coding assistant. All AI-generated content was reviewed, edited, and curated by the developer before inclusion. No AI is used to generate content live during gameplay.
Translations and text content are AI-generated. Some of the UI and game backgrounds are AI generated with manual post-processing. Some of the VFX for the animations are AI generated with post-processing.
Some of the text descriptions on the game’s store page, certain in-game localization texts, and some images used in the test version were generated with the help of AI technology.
The static pixel art images of the creatures (Deeplords) and environmental elements were pre-generated using AI tools and subsequently animated with another AI tool based on the developer's specific directions. Leveraging 10 years of professional 3D modeling experience, these assets were then integrated as textures and sprites inside complex 3D environments and dioramas modeled manually by the developer in Blender and structured in Unity. The game's Unity C# scripts were pre-generated using AI co-pilot tools based on the precise gameplay logic requested by the developer, followed by extensive manual debugging, correction, and multi-turn optimization cycles. Lastly, the soundtrack and audio elements were pre-generated using AI tools based on specific genre and stylistic vibes provided by the developer, then manually reviewed and refined to perfectly fit the dark atmosphere of the individual biomes.
This game uses generative artificial intelligence (AI) to create some of its visual art assets. Certain in-game images and illustrations were partially generated with the assistance of AI-based painting tools.
Kingdom of Ashes uses generative AI for select visual assets:
• Character portraits, enemy sprites, card illustrations, and background art were generated using Microsoft Copilot and then curated, composited, and refined by the developer for use in the game.
All gameplay code, mechanics, narrative text, and game design were created by the developer. No AI content is generated live during gameplay, and the game does not connect to any third-party AI service at runtime — all AI-generated visuals were finalized during development and ship as static assets.
Based on the developer's original creative direction, AI tools were used to generate the game's store page capsule images, soundtrack, and the non-English in-game text translations. AI coding assistance supported programming during development.
We're a small indie team, and we used generative AI to help make some of the art in Last Ember — card illustrations, backgrounds, character portraits, and a few short animations. We also used it on a handful of sound effects. Everything you see and hear in the game was picked, edited, and signed off by us. The game itself doesn't run any AI while you play — what ships is what you get.