Public Steam disclosure research25,394 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.
As an independent developer, I have used artificial intelligence to create some of the card images, as I do not have the necessary skills to make them myself; however, the plans for future releases include having enough capital for a real artist to create the graphics.
OxyRail contains a vast amount of content, with over 516 obtainable items currently available and many more planned for future updates. AI-generated content represents only a very small portion of the game. Due to the large number of items, we use AI-assisted tools in the creation of some item icons, certain background music tracks, and a very small number of visual assets to improve consistency and controllability during content production.
All AI-assisted content is carefully reviewed, adjusted, and further refined by hand to ensure it meets the artistic and design standards of the game.
We will continue to improve the game and gradually replace existing AI-assisted content with manually created assets whenever possible, increasing the proportion of original content over time.
Some visual assets used on the store page (such as promotional key art and background elements) were initially created with the assistance of generative AI tools and then manually refined and edited by our art team. No in-game content is generated using AI during gameplay.
Some early drafts of marketing texts were generated using AI and later reviewed and edited by a human.
Regarding the composition of game and store images, I used the AI model DeepSeek-R1 to generate them.As far as I know, images generated by AI usually do not have copyright
In Banana of Doom, AI takes the cards you play, pits them against each other, and generates funny, creative text reasons for the outcome then reveals them to the player
WilderLands uses AI to simulate NPC behavior and societies through reinforcement and deep learning.
AI agents adapt to the environment, form communities, make decisions, and learn over time.
They are not controlled by the player but evolve independently, creating personalized gameplay and unpredictable outcomes.
No AI was used for in-game graphics or music. This work is the product of human endeavor.
In the game, AI tools assisted with composing a singular audio effect that was otherwise not possible, as visual animation timing was fixed. Outside the game, AI was used to support marketing materials (circa 5% of marketing content).
There is no in-game generative AI. The clue generation runs locally, without a LLM or cloud service. We understand different people draw their own lines with AI. We aim to use modern tools responsibly, and hope this disclaimer provides for informed decision making.
The 3D models in the game are generated by Tencent Hengyuan 3D AI. Tencent officially stated: Service fee: To enable you to enjoy the 3D intelligent creation experience without any burden, this service is currently available for you to use for free. The introduction of Van Gogh's paintings was based on Baidu's Wenxin Yiyan. Wenxin Yiyan officially stated: The original copywriting generated by users based on Wenxin Yiyan belongs to the user themselves. It has been re-edited, modified, processed and polished by the author, and has undergone careful proofreading. The 12 large game scenes in the game were downloaded as paid scenes from the "Aigewe" website. I did indeed use artificial intelligence technology as an auxiliary solution.
Art: AI assisted with initial sketch concepts; final artwork is entirely hand-crafted.
Localization: AI translation used for initial processing; all text is human-reviewed and optimized.
Kayotica includes a unique AI-driven detective system. During certain scenes, players talk to the AI who takes a role of a characters. The AI is limited to the story world and cannot go outside the game’s narrative. Player choices in these moments influence relationships and the upcoming story choices.
Some images in the game are AI-generated but are consistent with our characters and our color chart.
Visual backgrounds and illustrations were partially generated using AI and then manually refined to meet artistic and technical requirements. Some voice lines were also generated using neural voice models, followed by manual editing and final polishing.
Some questions and answers in our quiz game are generated using AI. Additionally, we utilized pre-generated AI images included in the capsules and for in-game assets, specifically for the game characters.
Mini Scouts uses generative AI tools to create certain visual assets, including hero sprites, backgrounds, and select promotional images for the store page. All AI-generated content is pre-rendered, reviewed, and edited by a human being before inclusion in the game or marketing materials.
Now, we know that AI art cannot and should not compete with human art. Hence, we'd love to bring aboard interested human artists so we can replace our gen-AI assets with assets made with love by human beings!❤️
AI was used to create a reference base, enhance images, and generate some individual elements and objects. All outputs were then manually refined and polished by artists. The game also includes AI-generated video content; however, all characters depicted are unique and have been fully hand-polished by the artist to ensure originality and quality.
The game's store capsule image was created using AI tools. This is limited to promotional artwork only and does not affect the gameplay or in-game content.
As an independent developer, I have used artificial intelligence to create some of the card images, as I do not have the necessary skills to make them myself; however, the plans for future releases include having enough capital for a real artist to create the graphics.
OxyRail contains a vast amount of content, with over 516 obtainable items currently available and many more planned for future updates. AI-generated content represents only a very small portion of the game. Due to the large number of items, we use AI-assisted tools in the creation of some item icons, certain background music tracks, and a very small number of visual assets to improve consistency and controllability during content production.
All AI-assisted content is carefully reviewed, adjusted, and further refined by hand to ensure it meets the artistic and design standards of the game.
We will continue to improve the game and gradually replace existing AI-assisted content with manually created assets whenever possible, increasing the proportion of original content over time.
Some visual assets used on the store page (such as promotional key art and background elements) were initially created with the assistance of generative AI tools and then manually refined and edited by our art team. No in-game content is generated using AI during gameplay.
Some early drafts of marketing texts were generated using AI and later reviewed and edited by a human.
Regarding the composition of game and store images, I used the AI model DeepSeek-R1 to generate them.As far as I know, images generated by AI usually do not have copyright
In Banana of Doom, AI takes the cards you play, pits them against each other, and generates funny, creative text reasons for the outcome then reveals them to the player
WilderLands uses AI to simulate NPC behavior and societies through reinforcement and deep learning.
AI agents adapt to the environment, form communities, make decisions, and learn over time.
They are not controlled by the player but evolve independently, creating personalized gameplay and unpredictable outcomes.
No AI was used for in-game graphics or music. This work is the product of human endeavor.
In the game, AI tools assisted with composing a singular audio effect that was otherwise not possible, as visual animation timing was fixed. Outside the game, AI was used to support marketing materials (circa 5% of marketing content).
There is no in-game generative AI. The clue generation runs locally, without a LLM or cloud service. We understand different people draw their own lines with AI. We aim to use modern tools responsibly, and hope this disclaimer provides for informed decision making.
The 3D models in the game are generated by Tencent Hengyuan 3D AI. Tencent officially stated: Service fee: To enable you to enjoy the 3D intelligent creation experience without any burden, this service is currently available for you to use for free. The introduction of Van Gogh's paintings was based on Baidu's Wenxin Yiyan. Wenxin Yiyan officially stated: The original copywriting generated by users based on Wenxin Yiyan belongs to the user themselves. It has been re-edited, modified, processed and polished by the author, and has undergone careful proofreading. The 12 large game scenes in the game were downloaded as paid scenes from the "Aigewe" website. I did indeed use artificial intelligence technology as an auxiliary solution.
Art: AI assisted with initial sketch concepts; final artwork is entirely hand-crafted.
Localization: AI translation used for initial processing; all text is human-reviewed and optimized.
Kayotica includes a unique AI-driven detective system. During certain scenes, players talk to the AI who takes a role of a characters. The AI is limited to the story world and cannot go outside the game’s narrative. Player choices in these moments influence relationships and the upcoming story choices.
Some images in the game are AI-generated but are consistent with our characters and our color chart.
Visual backgrounds and illustrations were partially generated using AI and then manually refined to meet artistic and technical requirements. Some voice lines were also generated using neural voice models, followed by manual editing and final polishing.
Some questions and answers in our quiz game are generated using AI. Additionally, we utilized pre-generated AI images included in the capsules and for in-game assets, specifically for the game characters.
Mini Scouts uses generative AI tools to create certain visual assets, including hero sprites, backgrounds, and select promotional images for the store page. All AI-generated content is pre-rendered, reviewed, and edited by a human being before inclusion in the game or marketing materials.
Now, we know that AI art cannot and should not compete with human art. Hence, we'd love to bring aboard interested human artists so we can replace our gen-AI assets with assets made with love by human beings!❤️
AI was used to create a reference base, enhance images, and generate some individual elements and objects. All outputs were then manually refined and polished by artists. The game also includes AI-generated video content; however, all characters depicted are unique and have been fully hand-polished by the artist to ensure originality and quality.
The game's store capsule image was created using AI tools. This is limited to promotional artwork only and does not affect the gameplay or in-game content.