Where do games disclose AI use?

The game uses generative AI to generate store page images for the game, as well as Steam community assets and marketing materials and parts of the game UI

The game is a research project into how the use of AI would affect game production, so AI has been used for images, voiceovers, code, dialogue and other areas as much as possible. Just to see what sort of results it provides.

FACEMINER is a concept-driven game that directly engages with the ethics of AI. For thematic purposes, the game features faces which have been generated by a GAN (Generative Adversarial Network) trained on the FFHQ dataset, a research dataset released by NVIDIA that is the backbone of many real-world facial recognition applications. Per NVIDIA, only images under permissive licenses were collected. The 'AI training' in this game is fictional and has been simulated for entertainment purposes. The game can be interpreted as a critique of AI specifically relating to the ecological impact of mass data hoarding, surveillance concerns, and way it is used to compound corporate power.

Some backgrounds are taken from the Futuristic Anime Backgrounds Pack that is AI-generated.

AI-generated character illustrations.

Two profile pictures (one of these designs is also used as a decal on a tool), all of the "records" in the main characters room, designs on the vape packaging, and a background image for a vending machine were created with Adobe Generative Fill. The Security Robots and The Coffee Time video use audio generated by TTS Maker. The TOS was templated with AI as a generic TOS and meaningful elements were edited in. The Coffee Drink descriptions are pure, unfiltered, AI corpo-garbage-text. I used AI for these purposes because the uncanny-valley, soulless, style was exactly what I wanted as an art direction. I wanted to depict a future in which the soul of the world has been killed by AI. Making humans create art in the style of AI sounds completely dehumanizing, but you're entitled to your opinion.

Some voice-over lines are AI generated.

Prototypes for pregenerated content were used in one of the announcements.

Time Shock is hand-built by a solo developer. The design, gameplay, levels, writing, and world are all made by hand. AI tools play a limited supporting role only — used to tweak character voices, assist and refine artwork, and add lip-sync to dialogue. Everything is reviewed and finished by hand before it ships.

AI was used to assist with game localization, and create certain user interface elements.

This game uses artificial intelligence (AI) tools to help optimize code and improve development efficiency and runtime performance. Additionally, character voices in the game were recorded by the developer himself. The voice was then synthesised by AI to create different character voices. These technologies enhance the overall quality and experience of the game.

All voices for the announcer and the tutorial bot have been generated with AI

AI is used to create achievement icons.

AI assisted frame animation

The voices are generated using ElevenLabs and many of images are painted over and modified AI generations.

Both Generative AI and AI editing tools were used for some background and menu images, textures, and voices.

Used for Early Access placeholders, and is intended to be replaced with more development time. AI generation was used for some voices and some 2D images within the game. These assets are intended to be replaced with ones created by human artists.

Visual Assets: Some static assets, such as enemy hit reactions, were initially generated with AI-based tools and then edited by hand.

I added scenes and images improved with Generative Artificial Intelligence. They are locally generated using Stable diffusion, without resorting to paying any generative AI company. Although paying them could have given me better tools and results, I decided not to pay for their use to support the claim of artists for fair royalties. (It is not possible to discuss such a complex issue here, but I will consider opening a discussion on the topic in the game forum, follow my game to be notified if it is created) Most of the artistic content in the game, mainly music, is compensated in various ways depending on the case to its artists. My original content is registered, but the images generated by AI are essentially treated as 'creative commons' and I consider them fanarts of my characters. All added scenes and backgrounds are of high quality and maintain good coherence, avoiding deformities in hands, eyes, etc., and without annoying or confusing shapes. All AI-generated images are generated, selected, and edited exclusively by me, the creator of the game, to reflect my original vision, previously inaccessible due to high costs. I have sought to include only select content that enriches the narrative and immersion of the game.

The voice acting for some characters and some images were created using AI.

—Some background images and illustrations —Involving promotional art materials —A small number of image assets, including visual images and page backgrounds —Part of the content of the main promotional image

The audio for the voicelines for CARA (suit voice) and the ship announcer was generated by text to speech AI and then post-processed manually through an effect chain to sound immersive.

AI Voice generation was used to generate voice of the in-game AI character (AI assistant ZERA is speaking to the player). AI was used to help the human editor person to correct dialogues written by a human.

Deadeye Deepfake Simulacrum makes use of a custom built variational autoencoder to assist in the creation of character sprites, trained exclusively on the sprites I've already created. A separate generative adversarial network is used to make images that serve as the basis for some facial sprites, and background art.

As a solo developer I use AI assistance to generate graphics and audio content for my games and marketing material. No generative text is used.

The AI is used in one video in the steam and was used in videos between levels.

Generative AI (DALL·E 3) was used in Steam community resources. Specifically in trading cards and emoticon art.

Generative AI has been critical for the development of this game and is infact a part of it's lore as well. Generative AI assisted us in producing pre-renders for our backgrounds and complimenting the high quality of our human made animations, the backgrounds of this game wouldn't have looked as beautiful as it does nor even possible had BOS released before generative AI existed. In simple: The game uses AI generated images for the backgrounds of different regions.

In Ashes of Mankind, we utilize AI-generated voiceovers to bring life to key in-game characters, such as GRIM and E.M.A. These entities serve as guides and narrators within the battlefield, known as The Citadel. Their unique personalities and lore were crafted by our team, with their voices pre-rendered using AI during development. Importantly, this use of AI is not generated live during gameplay, and there is no intention to impersonate real individuals or infringe upon existing copyrights. These AI-generated voices are carefully designed to enhance immersion while respecting all applicable guidelines. We remain committed to creating an engaging and authentic experience while ensuring transparency and compliance with Steam's policies.