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 game includes an unholy amount of AI-generated content, including:
- Textures
- Images
- Music
- Audio effects
- Voice lines
- Facial and body animations
- Code
Game allows player to use voice as a main input tool. To achieve great quality of speech to text, a base version of openai's whisper model is used.
The content seen in the game is not live-generated. The player's input is embedded into a vector through a small model with 278M parameters.
Model:
paraphrase-multilingual-mpnet-base-v2
(Used primarily for wide support of languages)
Then, the game uses Faiss to match the player's input to known, predefined inputs. Each of those describes a reaction and change in the state of the game.
Icon was made with local diffusion model.
In the game, artificial intelligence is used only in a few aspects, specifically for generating voice acting and music. All the music featured in the game, as well as the voice acting for characters, is entirely created using AI technologies.
I use the service to generate artificial intelligence voices for characters. The voice generation process does not occur in real-time, but rather beforehand, allowing me to test and evaluate the content before using it in the game. Additionally, I implement appropriate preventive measures to ensure that the generated content is suitable and compliant with legal requirements and Steam standards.
The game uses some AI generated images for some of the products in our game, they are also used for some set dressing in the game world. These images are heavily modified after they are generated to match the style of the game but we want to have full disclosure that we used AI imagery in different places in the game and that we intend to replace if not all most of them by the time we have a full release.
Some graphic content on the store page was generated with Artificial Intelligence. In the game, some graphic images, the paintings hanging around the house, and the voice acting are created with Artificial Intelligence.
Some of the posters were created using generative AI (Adobe Firefly).
All of the music tracks were created using generative AI (Suno AI).
The translations were created using generative AI (ChatGPT, DeepL Translate).
Our game utilizes AI in several ways. We generated some concept art using AI, which helped shape the visual direction of the game. Additionally, we partially used AI for translation tasks to ensure a broader audience can enjoy the game. AI also assisted us in some coding tasks, streamlining development and allowing us to focus on crafting a better gameplay experience.
AI technology is used in the creation of 3D models.
Logos and backgrounds used on the store page are generated using AI.
AI is utilized when searching for sample music materials.
Some of the in-game text was created with the assistance of AI.
Most of the images used in the game are generated using AI.
There is a radio station that put on air AI generated songs, random generated NPCs have access database of pre-generated topics of discussions, there are collectable lost media VHS featuring short experimental films made with AI generated and real footage
Occasionally on real-life blink in-game textures may swap to pre-rendered AI variants of the textures for a moment.
These textures were created with an outdated image generation model on local hardware to facilitate hallucinations. There was no new training of models for SWEYET(etc.), no wastage of water, etc. No generative AI runs on users machines, there is no runtime generation.
Throughout development, I may use procedural methods and AI-based tools to assist with audio content creation (10%). However, the final result always reflects my own creativity and artistic intent.
The game includes an unholy amount of AI-generated content, including:
- Textures
- Images
- Music
- Audio effects
- Voice lines
- Facial and body animations
- Code
Game allows player to use voice as a main input tool. To achieve great quality of speech to text, a base version of openai's whisper model is used.
The content seen in the game is not live-generated. The player's input is embedded into a vector through a small model with 278M parameters.
Model:
paraphrase-multilingual-mpnet-base-v2
(Used primarily for wide support of languages)
Then, the game uses Faiss to match the player's input to known, predefined inputs. Each of those describes a reaction and change in the state of the game.
Icon was made with local diffusion model.
In the game, artificial intelligence is used only in a few aspects, specifically for generating voice acting and music. All the music featured in the game, as well as the voice acting for characters, is entirely created using AI technologies.
I use the service to generate artificial intelligence voices for characters. The voice generation process does not occur in real-time, but rather beforehand, allowing me to test and evaluate the content before using it in the game. Additionally, I implement appropriate preventive measures to ensure that the generated content is suitable and compliant with legal requirements and Steam standards.
The game uses some AI generated images for some of the products in our game, they are also used for some set dressing in the game world. These images are heavily modified after they are generated to match the style of the game but we want to have full disclosure that we used AI imagery in different places in the game and that we intend to replace if not all most of them by the time we have a full release.
Some graphic content on the store page was generated with Artificial Intelligence. In the game, some graphic images, the paintings hanging around the house, and the voice acting are created with Artificial Intelligence.
Some of the posters were created using generative AI (Adobe Firefly).
All of the music tracks were created using generative AI (Suno AI).
The translations were created using generative AI (ChatGPT, DeepL Translate).
Our game utilizes AI in several ways. We generated some concept art using AI, which helped shape the visual direction of the game. Additionally, we partially used AI for translation tasks to ensure a broader audience can enjoy the game. AI also assisted us in some coding tasks, streamlining development and allowing us to focus on crafting a better gameplay experience.
AI technology is used in the creation of 3D models.
Logos and backgrounds used on the store page are generated using AI.
AI is utilized when searching for sample music materials.
Some of the in-game text was created with the assistance of AI.
Most of the images used in the game are generated using AI.
There is a radio station that put on air AI generated songs, random generated NPCs have access database of pre-generated topics of discussions, there are collectable lost media VHS featuring short experimental films made with AI generated and real footage
Occasionally on real-life blink in-game textures may swap to pre-rendered AI variants of the textures for a moment.
These textures were created with an outdated image generation model on local hardware to facilitate hallucinations. There was no new training of models for SWEYET(etc.), no wastage of water, etc. No generative AI runs on users machines, there is no runtime generation.
Throughout development, I may use procedural methods and AI-based tools to assist with audio content creation (10%). However, the final result always reflects my own creativity and artistic intent.
