Where do games disclose AI use?

Some visual assets — such as community materials, marketing content, and select transition screens — were created with limited AI assistance. All final in-game art is created by our artists.

GlitchSPANKR was handcrafted. We used AI audio sparingly and purposefully: We remixed our own original tracks into ambient variations (e.g. a French-style version of “Friend”). A handful of songs had lyrics we wrote and used AI to help generate their music, which we refined and often re-recorded ourselves. One minor joke character uses an AI voice by design.

At the current stage the game contains AI-generated voice acting and a single AI-generated image asset. This *may* change in the finished version.

GitHub Copilot for coding. 2d UI Icons. External promotional art - not inside the game.

Part of the screen is assisted by AI

In Rivandy, Artificial Intelligence is only used as a creative assistant. Since we are a very small team, AI helps us mainly with artistic aspects—such as illustrations to enhance immersion, or voice generation to ensure consistency across characters. Occasionally, it may assist in code prototyping to unblock situations. However, the game itself is essentially human-made: the core code, design, and mechanics are developed by us, and all visual assets such as FX and 3D models are purchased, not AI-generated. This way, we can stay focused on building a unique, handcrafted experience while using AI only as a supportive tool.

AI used for some of the translations.

Generative AI was used as a step in the development of some texture and model assets.

Voice-over for NPCs.

AI was used to assist in generating art

We used some AI-tools to help develop the game.

Used for creating the game banner

AI is used for parts of the English localization and for localization into other languages.

Artificial intelligence was used to create the images.

We have used AI generated artwork for our store listing, certain in-game art assets have been AI generated however majority of the game is not AI generated.

For Electric Alps, we utilize generative AI to aid in creating background assets and promotional materials, achieving a quality that would otherwise not be feasible for our small studio. All gameplay elements are meticulously handcrafted by our artist for a unique playing experience.

This game makes use of state of the art generative text and voice services. Shouldermen relies on OpenAI GPT-5 for dialog and Eleven Labs for custom voices and Deepgram for voice input.

Some textures in Candy Bandits were initially generated using AI tools and then manually edited and adjusted by our artists to meet our design and quality standards. No generative AI is used during gameplay or player interaction.

Use of AI for visual materials on the Steam page due to limited resources and budget. This is done to try to give the page a better appearance than what we could achieve with our own skills.

some pictures and charaters are made by AI.

All still images in the game are generated by AI

The game's intro video and ingame backgrounds were first created with 3D renders and then styled using Stable Diffusion. However, no images were created from scratch with ai. The resulting visuals were then cleaned up and refined using PixageFX and a pixel art editor. Additionally, a large portion of the character voiceovers were done using 'murf.ai' except 'Mike Jack'.

Partial image assets are generated using AI assistance

Some card & store page artwork are ai assisted.

We use AI tools to help us develop the game faster, mainly for coding purposes.

Character and hero images are generated as placeholders and we are in the process of hiring artists to replace the art. 100% of the game's profit goes to replacing AI art. Currently >50% of art is already replaced (non-AI).

To make references of previously made designs, to later do post production of them

Some 2D artworks and promotional images in our game are created using generative artificial intelligence.

We've created voice lines for the player's hub A.I. companion, known as L.A.N.A. (short for Logical Artificial Neural Assistant), to narrate tutorial texts. They are all pre-rendered.

AI was used to generate some of the images and to animate the historical characters.

For recipe art only: Our artist used an AI assistant. We made a hand-painted recipe base. Then we used AI to add a detail layer. Then we hand-painted all final details.

Some voice acting is generated using AI models trained on professional voice actors who have given their consent and are being compensated for their work.

AI tools have been like an extra set of hands (or brains) for our small indie dev team. They've played a quiet yet invaluable role in helping us draft initial concept art and brainstorm our way through tricky design challenges. Being Polish native speakers, AI has also been particularly helpful in fine-tuning our writing, ensuring our game's narrative is as engaging as possible.

Just for some image like the wizard for example Also was used to translate the game on various lenguages

Some models and icons in the game are generated by artificial intelligence

Wacky West had some AI art placeholders that were since replaced with hand made assets.

The game uses real-time AI models to simulate immersive conversations with the non-player characters you meet in-game. Each character has a hand-written backstory that impacts how the conversation goes. This technology complements the manual worldbuilding and lore writing, and I would love your feedback on how well you think this works!

This game is fully human‑programmed. All objects, characters and enemies were hand‑drawn with only one exception: The Rex head in level 10. A few AI‑generated images were used for Steam and simple backgrounds.

Characters and graphics made with Midjourney. Title theme made with Aiva.

Character portraits were made using AI. The two paintings in the main menu in the background are also AI generated. Intro movie contains AI images.

Generative AI (DALL·E 3) was used in Steam community resources. Specifically in achievements art.

Some in-game content, such as simulated chat spam and cutscene art in WISH (as befits the narrative) was created with the assistance of AI powered tools.

Some of the store artwork (Steam capsules) are designed with assistance of generative AI. The game itself - includding the music, levels, code and all artwork - is "hand made" without any of this AI stuff.

Some translations are AI-generated.

As an independent developer, I use AI during development as a learning and workflow assistant. It helps me understand tools, write a small portion of code, and polish some localization. All final content in the game is personally reviewed and tested.

Angel Guardian was developed with the help of AI tools across our creative pipeline — including story development, dialogue, enemy balancing, difficulty tuning, concept art exploration, animation prototyping, music iteration, and voice experimentation. AI acted as a creative accelerator, allowing us to rapidly test ideas, refine systems, and expand the scope of our world. All content was directed, curated, and finalized by our human development team.

Backgrounds

The cover art and character poses and expressions in our game were initially generated using AI during the development phase to enhance the creative process. These elements were then redrawn by our art team to ensure quality and originality. Additionally, the story text within the game was refined, proofread, and translated with the assistance of AI tools. No real-time AI content is generated during gameplay, ensuring players receive a consistent and controlled gaming experience.

For the MVP of the Development version of the game 3D models, 2D sprites, sound and music are part wise AI generated. These assets will not be used for the release of the game but instead replaced by human-made assets.

Narrator uses AI voice over work as a placeholder. Currently the final game is slotted to feature 100% human actors for maximum expression. An example of this is the demo's ending cutscene.