Where do games disclose AI use?

本游戏在部分功能中使用人工智能技术辅助内容生成与体验优化，所有AI生成内容均由开发团队预先审核、筛选与管控，不会自动生成实时、无审核的动态内容。 游戏内AI相关功能主要用于优化游戏体验，包括但不限于场景细节渲染、剧情文本润色、智能数值适配、互动内容辅助生成等。所有AI产出内容均严格遵守内容合规规范，不含暴力、色情、违法、歧视、虚假误导等不良内容，同时不会收集、上传用户个人隐私信息。 本游戏所有AI相关内容均为游戏内置固定功能，无用户自主输入 prompt 自由生成内容的玩法，不存在无限制AI生成内容的风险，所有展示内容均经过人工审核与版本固化，保证内容安全、可控、合规。

This game contains generative AI content. Some visual/visual arts materials were assisted by generative AI tools during development and were manually selected, edited, and reviewed by the development team before being used in the game or Steam store page. The game features optional real-time AI chat and voice interaction functions. Players can interact with in-game pets through text or voice input, and the backend AI generates text responses based on player input, current game state, and context. Some responses may be delivered via voice. We employ input filtering, output filtering, safety prompts, rate limiting, and abuse detection to minimize the generation of illegal, infringing, explicit, hateful, harassing, or otherwise inappropriate content.

In this work, much of the text is interactively generated by a language generation AI. In addition, language generation AI is used for various judgments on the text entered by players. Furthermore, all the illustrations in this work are created using image generation AI.

SonderSoft uses generative AI for localization to make Arena Bot Manager accessible to more players worldwide. ABM features an optional AI voice-commentator for arena events that can be disabled as per player preference. SonderSoft pays artists to participate in the creative process, and some artists may choose on their own to use generative AI as part of their workflow; SonderSoft respects the creative tools artists choose in order to realize their own artistic vision. Whenever possible, SonderSoft prioritizes paying artists, creators, and actors.

The application uses AI to provide real-time translation functionality.

GenderShift uses generative AI in two ways: 1. Live audio: AI converts your speech or text into the voice you select 2. Voice portraits and sample audio (in the app, trailer, and screenshots) were pre-created with generative AI. You can also generate these yourself in the app.

SWIFTALE PLAYER uses generative AI for live text generation during gameplay. The player connects their own supported local or cloud LLM, and the model acts as a game master by generating narration, action suggestions, and story responses. The game does not include a hosted AI service. AI output is processed through structured game systems, including rule-based safety checks, resource/state management, and engine-side adjudication constraints. The intended content profile is general fantasy adventure, and the game is not designed to generate sexual content, hate content, extreme gore, or adult-only material.

EchoMind uses a locally running large language model to generate conversational text responses in real time based on the user’s prompts and local memory/context. The AI-generated content is limited to text-based conversation and local memory summaries. EchoMind does not generate images, music, voices, sound effects, 3D models, textures, or executable code as part of its normal user experience.

This game uses generative artificial intelligence to create dynamic content during gameplay. AI systems may generate political advice, news articles, speeches, social media posts, events, narrative responses, and other text-based content based on player actions and the current game state. AI-generated content is intended to enhance replayability and create unique gameplay experiences. Some store, library, and character artwork was created with the assistance of generative AI image tools.

Command Doctrine uses generative AI to create some of the content you play. When you generate a scenario or a campaign, an AI model produces the map, the opposing forces, the objectives, and the written mission briefing from your prompt. AI is also used to write narrative battle reports, to create the images that accompany them, and to interpret the natural-language orders you give your units. This content is generated live by third-party AI services. We apply content moderation and input filtering to guard against inappropriate or unsafe output.

We use a local real-time deep learning neural network (AI) to perform depth estimation on the video frame. The AI analyzes the 2D image and generates a corresponding depth map in real-time, which is then used to reconstruct the 3D VR environment. The AI only processes the visual depth data and does not generate any text, audio, new art assets, or user-facing media content.

This game uses AI-assisted tools during development, including assistance with coding and the creation or iteration of some promotional and visual art assets. During gameplay, AI may be used to help interpret the player’s spoken spell commands and map them to predefined in-game spell actions.

The game calls the aforementioned third-party large language model API over the network at runtime. When the player makes a decision, issues an imperial edict, or interacts with ministers in the game, the game assembles the current situation and character settings into a request and sends it to the AI service; the text returned by the service (minister dialogue, memorials to the throne, plot deduction) is presented in the game interface in real time. This process is transparent to the player and requires no player-side configuration, but an internet connection is required throughout.

Netshell uses AI to power its characters. The NPCs you meet on the in-game forums and IRC read what is happening in your playthrough and respond in character, so their messages and forum posts are generated rather than pre-scripted, and each player's world unfolds differently. All of this runs locally and offline: the AI models ship with the game and run on your machine. Netshell never connects to a third-party AI service such as ChatGPT or Claude. You can turn the AI off entirely, in which case the NPCs stop chatting and the game continues with its authored story. Some Steam store and trailer assets were created with AI assistance and then edited/curated by the development team. During gameplay, Netshell only uses AI to generate in-character NPC text.

Penko Vox: Japanese uses local generative AI to power its immersive language learning experience. An integrated Large Language Model (LLM) acts as a native Japanese tutor, generating real-time dialogue based on user input, while a Neural Text-to-Speech (TTS) engine synthesizes high-fidelity Japanese audio for listening practice.

IKANDY includes an optional AI scene generator. When you choose to use it, it creates new audio-reactive visual scenes in real time from a text prompt you write. This content is generated live, only when you request it, using an AI service you connect with your own API key. Not required to use IKANDY. Some of IKANDY's artwork and promotional media (key art, trailer visuals and music, and some in-app imagery) was created with generative AI tools and is reviewed by the developer.

The NPCs in the game integrate AI decision-making and dialogue functions.

In this game, questions and answers (yes/no) are sent to Amazon Bedrock to generate in-game text. Personal information like your name is not sent. Progress data is saved on your device.

English (AI Disclosure) This game includes AI-assisted NPC dialogue and text generation to create dynamic conversations and world events. AI-generated text is reviewed and integrated by the developer before release. The game does not generate images, audio, or code on the player's device. AI is used to enhance dialogue variety and immersive world interactions while core gameplay, progression, and game systems are designed by the developer.

This game lets you converse with a character based on the historical Japanese warlord Oda Nobunaga. The character's reply text is generated on the fly by a language AI that runs locally on your own PC in response to your input. ・Everything runs entirely on your PC. No data is sent to any external server, and the game works without an internet connection. ・A system prompt (character definition) keeps the AI in role: speaking in Nobunaga's style and offering encouragement. It answers based on the historical record and will honestly say it doesn't know when it doesn't. ・The only thing generated at runtime is conversation text. The portrait uses a public-domain painting, and the talking and blinking animations based on that portrait are fixed clips created with a video-generation AI during development and played on a loop. Speech is produced by a voice-synthesis engine (VOICEVOX). ・This is a conversational app made for encouragement and light counsel. The AI's responses vary every time and are not guaranteed to be accurate. Please enjoy it as entertainment.

This game includes live-generated AI features that run locally on the player's computer and do not rely on external AI services. The game uses AI to: - interpret the player's spoken voice commands in real time to support gameplay interactions; - generate voice audio for in-game pilot communications in real time. The AI systems are used exclusively for these gameplay functions. They do not generate user-visible images, text, or other open-ended content. The game is designed so that AI-generated output is limited to its intended gameplay purpose and cannot be used to create illegal, infringing, or otherwise prohibited content.

This application uses generative AI models (Depth Anything V2 / V3) to perform real-time monocular depth estimation on user-provided videos, converting 2D footage into 3D depth maps for stereoscopic and immersive viewing. All AI models are loaded and run locally on the user's own hardware. No third-party cloud AI services are used, and no user data is transmitted externally.

Arcadia Unbound does not use AI-generated artwork, music, sound effects, voice acting, story content, quests, localization, or character dialogue. Arcadia Unbound uses a locally running language model as part of the AURA Framework to interpret player speech during specific conversation sequences. The model analyzes player intent and conversational context, and then selects from developer-authored dialogue responses and narrative outcomes. All artwork, music, voice acting, story content, character dialogue, and gameplay content presented to players are created and curated by the development team and performed by professional voice actors where applicable. No AI system generates any type of dialogue, story events, artwork, audio, or player-facing content.

Game Overlay Translator uses AI at runtime to translate the on-screen text that the player chooses to capture from other games or apps. All processing runs locally on the player's own PC and works offline: on-screen text is recognized with an on-device OCR model (RapidOCR / ONNX Runtime) and translated with on-device neural machine translation models (FuguMT, Argos Translate) and, optionally, a locally-run LLM (Ollama / Qwen2.5, or NVIDIA Riva). No captured text, images, screenshots, or any other user data are sent to external or cloud AI services. The app does not accept free-form prompts and does not generate original or creative content — the AI only outputs a translation of the specific text the player captures. The source text belongs to the third-party games or apps the player runs. Guardrails: the tool translates only the text the player explicitly captures from their own screen; it cannot be prompted to produce anything else, and all inference is local/offline, so no content is generated on, or sent to, any external server.

This game uses AI to assist in proper procedural generating level

This product uses generative AI for live text-based character dialogue, optional code or technical text generation, and AI-assisted pre-generated visual assets. Players can interact with an AI desktop character through text chat. AI responses may be generated based on the character profile, the current conversation, limited memory, and optional screen-context summaries. When requested by the user, the AI may also generate programming explanations, code snippets, configuration examples, or technical suggestions. AI conversation features require an internet connection. In the Steam version, players do not need to create, configure, or manage any third-party AI service accounts or API keys. Access to AI features is managed by the developer’s backend service and uses in-app AI Credits purchased through Steam Wallet. The artwork used as the basis for the built-in character was created with assistance from ChatGPT and was subsequently manually edited, separated, and rigged by the developer as a Live2D model. Some promotional images, Steam library assets, background art, and key art were also created or refined with the assistance of generative AI tools during development. The product does not generate images, audio, music, voice, textures, 3D models, or video at runtime. Generated code is not automatically executed, compiled, installed, or applied by the application. Any use of generated code is controlled by the user outside the application.

MultiScope Air Traffic Control uses a local neural text-to-speech model (Piper) to generate the pilot voice read-backs in real time on your own machine. It runs entirely offline, no voice input or player data is sent anywhere or used to train any AI model. Your spoken commands are recognized locally by your operating system, not by generative AI.

The game uses artificial intelligence to control characters behavior and deliver an adaptive fairytale plot. The models, sounds, maps, and other assets are created by real people, with love and care.

VoxelForge includes local generative AI features for user-assisted 3D and voxel asset creation. Users can provide their own source images and use the application to generate 3D models, voxel assets, meshes, textures, and related production assets. This generation runs locally on the user’s computer using model files bundled with the application. VoxelForge does not connect to an external AI service during normal use. It does not require a Hugging Face account, API key, cloud generation service, or external server connection at runtime. User input images and generated outputs are not uploaded to our servers. The AI features are intended as creative tools for asset production. Users are responsible for ensuring that they have the necessary rights to any images or source material they import, and for reviewing generated outputs before using or distributing them.

Players can freely converse with NPCs, and the game generates dialogue in real time based on the NPCs’ personalities and memories. The main storyline will be handled by AI agents, which will generate plots consistent with the game’s world-building. AI assists in creating CG scenes, background images, and character portraits.

We have carefully crafted a complete set of NPC dialogue without the use of AI. At the same time, we are experimenting with an additional feature powered by AI large language models, allowing players to freely converse with their favorite NPCs. This is an exploratory feature that exists entirely separately from the traditional game content. These experimental elements are intended to offer players a new additional experience and will not affect the complete single-player experience. Whether this feature will be included in the final release has not yet been decided. Even if it is ultimately implemented, it will be optional rather than required.

Silver Creek Town uses generative AI to support live NPC dialogue text during gameplay. Player text input may be processed by an AI service to generate in-character fictional NPC responses within the world of Silver Creek Town. Some background music, concept art, character images, text drafts, and marketing materials may also be created or assisted by AI tools during development. We review AI-assisted content before use and apply prompts, content rules, and safety restrictions to reduce inappropriate, illegal, hateful, sexually explicit, infringing, or unsafe output.

This game uses generative AI in two ways: (1) AI language models (LLM) to generate character dialogue text in real-time during conversations, and (2) AI text-to-speech (TTS) technology to generate character voice audio. AI processing is handled via cloud servers or optional local models. No conversation data is stored or saved on servers. All AI-generated content is produced live based on player input.

How generative AI is used in this game: 1. Pre-generated Content This game was fully developed without manual human intervention using generative AI tools. Code & Development: The entire codebase, programming logic, user interface, and system architecture were 100% written by AI coding assistants. Design & Assets: All visual elements, text, menus, and UI components were pre-generated using generative AI. 2. Live-generated Content The core gameplay revolves around an interactive AI companion ("AI Buddy"). Real-time Interaction: All dialogues, responses, personality traits, and interactions of the AI Buddy are generated live and dynamically during gameplay using a Large Language Model (LLM). Safety & Guardrails: The game utilizes integrated safety filters and guardrails provided by the AI model. This strictly ensures that the live-generated content does not produce any illegal, hateful, copyrighted, or inappropriate material.

The AI acts as your gameplay partner (either as a captain or a guesser). It analyzes the words on the game board and generates short text clues, or tries to guess the correct words based on the text clues you provide.

Popular Kids uses generative artificial intelligence to power conversations with student characters. Players can interact using voice or text, and AI-generated dialogue allows characters to respond dynamically based on the current situation, their personality, previous interactions, and the player’s choices throughout the game. All storylines, characters, progression systems, popularity mechanics, reputation systems, and overall game design are created by the developer. Generative AI is used only to generate character dialogue and conversational responses during gameplay. AI-generated content is monitored and guided by character-specific instructions and game rules to maintain age-appropriate school-life interactions and narrative consistency.

Playstories uses live-generated AI during gameplay to create fictional interactive story prose, dialogue, player choices, scene keywords, branching follow-up scenes, and optional scene images from player prompts and actions through novelai.ai. AI does not control purchases, account actions, executable code, or system-level behavior.

This game uses AI language models to generate real-time conversational text content during gameplay. Players interact with AI-powered desktop companions through text chat. The AI responses are generated either locally on the player's PC or through optional cloud AI services (such as OpenAI, Deepseek, etc.). All AI-generated content consists of text-based conversations only — no AI-generated images, audio, or video are used.

Hire uses AI for its live character conversations and generated portraits, connecting to your own ChatGPT account (it uses your existing ChatGPT plan, with no separate purchase).

The game includes generative AI in the form of relatively structured conversation (voice and/or text to prompt) in context with the setting and era of the documented sites (caravanserai) you will visit.

VoidCraft uses generative AI in two ways. During development, AI tools assisted in creating written content that ships with the game (lore, item and research descriptions, mission templates, NPC flavour text) and in exploring marketing artwork concepts; all of it was reviewed by the developer before inclusion. During gameplay, two features generate text live using large language models: the optional in-game Advisor, which analyses your empire and offers strategy advice in character, and contract mission briefings (titles, short stories, and NPC names) on station contract boards. Players cannot type free-form prompts into these systems - the AI receives only structured game-state data assembled by our servers, and every output is validated server-side before it appears in the game. Core gameplay (combat, economy, progression, rewards) is governed by deterministic game rules, not by AI.

About Art Assets DesktopDaily does not use AI-generated artwork for its in-game art assets. Character designs, 3D models, textures, animations, UI assets, and other visual content are created by human artists or based on properly licensed materials. About Character Voices and Speech Synthesis DesktopDaily’s character voice models are trained or fine-tuned using voice materials with clear provenance and authorization, and are used for character speech synthesis at runtime. These voice materials come from publicly available or authorized audio materials provided by the corresponding character or voice provider, and are used in accordance with their usage terms and authorized scope. The character voices played at runtime are synthesized speech and are not original voice recordings. About the Use of Generative AI During development, AI assistance was used for parts of the code implementation and for some non-Chinese UI text localization. At runtime, generative AI is used only in optional features, including the character conversation feature, the visual companion feature, and notification assistant tools. Some non-Chinese UI text localization uses AI-assisted translation and editing. This is limited to UI text localization and is separate from the runtime generation of character dialogue. In the optional character conversation feature, AI is used to generate character dialogue text and to synthesize speech based on that text. When this feature is enabled, player input may be sent to cloud AI services in order to generate character-style response text. These responses may also be used for speech synthesis and played back in the game. If the player chooses not to use an API, the character conversation feature will not be enabled, and only the basic features will be available. In the optional visual companion feature, after the player actively enables the feature, the game may use a local vision-language model to analyze screen, window, or application content within the scope authorized by the player. The analysis results may be used as context for character dialogue, allowing the character to respond based on the player’s current computer usage scenario. This feature is disabled by default. If the analysis results are used in cloud-based character dialogue, necessary summary information may be sent as part of the dialogue context to the cloud AI service selected or configured by the player. In the optional notification assistant tools, after the player actively enables the feature and grants permission, the game may read Windows system notifications, email summaries, or other information sources specified by the player. A local AI model may be used to filter, categorize, or judge the importance of this information in order to decide whether to notify the player. This feature is intended to assist with notification management, and will not reply to emails, send messages, or perform external actions on behalf of the player. AI-based classification and notification results may be inaccurate. The official trial API provided by the game uses Alibaba Cloud Model Studio. Players may also configure other APIs themselves. When features that require cloud AI are enabled, the necessary conversation content or contextual information is sent directly to the corresponding cloud AI service and is not relayed through the developer’s own server. Character dialogue generated by AI at runtime is intended only for everyday companion-style interaction. The game uses prompts, content rules, and safety measures to reduce the output of inappropriate, illegal, hateful, sexual, violent, or copyright-infringing content. AI-generated responses may vary and do not represent official statements by DesktopDaily, the game’s developer, any character rights holder, voice provider, or third-party asset creator.

The game includes AI-driven enemy characters ("LLM zombies") that players can interact with through dialogue. These characters use a language model to generate responses dynamically during gameplay. The AI is used only to produce in-game dialogue and behavior for these enemies and does not generate user-uploaded content or external assets.

This game contains real-time AI-generated text.

This game uses generative artificial intelligence during gameplay to generate text and data based on player input and game context. The game does not use generative AI to generate artwork, audio, voice, music, videos, or other visual or sound assets. AI-generated content is limited to text and data used as part of the gameplay experience.

AI-driven real-time asset generation is a core part of the player experience in Project Resonance. Some of the game's in-game assets are generated in real time during play using generative AI—including music (melody, arrangement, and vocals), character art, expression and sprite variations, live-performance visuals, event illustrations, event storylines, and roleplay dialogue. Because this content is generated in response to your choices and progress, every playthrough produces music, visuals, and story unique to that session. All AI-generated content is screened by our in-game filtering and review systems at runtime—before it reaches the player—to prevent inappropriate, infringing, or off-tone output and to keep everything consistent with the game's intended tone and content rating.

This game uses generative AI to create realistic player behavior, match commentary, dynamic tactics, and some images and music may also be generated by AI, improving replayability while keeping user data private and not used to train external models.

Thirsty Snail uses an on-device local language model to generate short snail dialogue, personality-based reactions, and race taunts. AI-generated text is triggered by in-game interactions and constrained by character prompts and content moderation rules. Screen-aware reactions are processed locally on the player's device. Screenshots are never stored or uploaded, and private file contents are not read.

dreamstrobe uses generative AI as a core live feature. When you chat with the visualizer, your prompt is sent to Anthropic's Claude model. Claude writes a Three.js scene — geometry, shaders, motion, post-processing, and audio-reactive behavior — in real time, and the running engine loads and renders it on the fly. Each scene you create is freshly, generated JavaScript that runs in a sandboxed browser context. Outputs are filtered by Anthropic's built-in safety systems to block illegal or harmful content. Generated code is executed locally; no model weights ship with the game. Some textures, environment maps, and sample assets used by the engine were also created or refined with generative AI tools during development, and were reviewed before inclusion. You can use dreamstrobe without writing prompts — a library of pre-authored scenes ships with the game and runs on its own.

This game uses generative AI to produce dialogue content in real time. All character responses are generated by a large language model (LLM) based on user input and the loaded character card settings. No pre-written dialogue scripts are included in the shipped game. Character cards are created by third-party authors and may themselves contain AI-generated content.