AI Use Policy
Last revised 2026-10-05
Effective date: October 5, 2026
1. Ddalkkak is a generative AI service
The script, voice, scene images and video that Ddalkkak makes are made by generative artificial intelligence (AI). Each step calls outside AI services, and this document sets out what goes where, and how results should be handled.
2. Which companies’ AI we call
| Company | What it does | Country |
|---|---|---|
| OpenAI | Writing scripts, scene images, generating video descriptions and thumbnails | United States |
| Google LLC (Gemini·Veo) | Reference analysis, script and topic ideas, generating scene images and video | United States |
| Anthropic | Writing Shorts scripts, analyzing the visual style of reference videos, reviewing generated images | United States |
| ElevenLabs | Generating voice (narration) | United States |
| xAI | Generating scene images and video (when that model is chosen) | United States |
| fal.ai (Features and Labels) | Generating scene images and video — relaying the Seedream, FLUX, Kling, Seedance and Wan models (when that model is chosen) | United States |
| MiniMax | Generating scene video (when that model is chosen) | Singapore |
This list contains only what is actually in use at present, and it is the same list as the overseas transfer table in the Privacy Policy. Website hosting (Netlify) and connection relay (Cloudflare) are not AI, so they are not in this table; they are listed in the Privacy Policy. If the services we use change, we revise this document first and then apply the change.
3. What is transmitted
- The topic sentence and script you entered and the reference video address
- Captions and a few frames read from that video
- The script, scene descriptions, images and voice generated at each step
Account information (email, name, password) is not sent to outside AI services. Only the content needed for production is sent.
4. Rights and responsibility for results
- The right to use the generated script, images, voice and video belongs to the user.
- However, results made by AI may by chance resemble existing works. Please check for yourself before publication or commercial use. We do not warrant that outcome.
- A reference video is used to reference its visual grammar and narrative structure, and the service is designed not to copy the original’s story or scenes. However, it cannot prevent cases where a user enters input to follow the original as it is. Confirming rights in the reference video is the user’s responsibility.
5. Please disclose that it was made with AI
- When you publish a result, please disclose that it was made with AI. When uploading to YouTube, you can use the “Altered or synthetic content” item on the upload screen.
- In particular, for results that look like real people, events or places, state it clearly in the video or in the description so that viewers do not mistake them for reality.
- Some AI companies also embed their own invisible marks in the images and videos they generate. Ddalkkak does not add a separate mark to results, so labeling at the time of publication must be done by the user.
6. What a person must check
AI can get facts wrong. Do not take numbers, dates, quotations or statements about people at face value; please check them before publication. That is why each step has a screen where you view and confirm the result.
7. Is it used for training
We do not use users’ inputs and results for model training. The data handling of outside AI services follows each company’s policy — links to each company’s policy are in the overseas transfer table of the Privacy Policy.
8. Cost
Generation steps trigger calls to outside services and use credits. The screen shows the estimated credits before you start, and no call is made before you press the start button. In demo mode, saved example results are shown and no outside calls are made.
9. Contact
For inquiries about this policy, please write to official@aigorhythm.net.
View previous AI Use Policy (September 20, 2026)
Ddalkkak calls outside AI services at each step of video production. This document sets out what goes where, and how results should be handled.
1. What is called at which step
| Step | What it does | Sent to |
|---|---|---|
| Reference analysis | Reads the captions and frames of a public YouTube video | YouTube (Google) |
| Script | Writes the script from the topic and the analysis result | OpenAI |
| Voice | Generates the voice that reads the script | ElevenLabs |
| Scene images | Draws the character and background sheets and the scene images | OpenAI |
| Shorts scene video | Turns images into moving clips | MiniMax |
This list contains only what is actually in use at present. If the services we use change, we revise this document first and then apply the change.
2. What is transmitted
- The topic sentence you entered and the reference video address
- Captions and a few frames read from that video
- The script and scene descriptions generated at each step
Account information (email, name, password) is not sent to outside AI services. Only the content needed for production is sent.
3. Rights and responsibility for results
- The right to use the generated script, images, voice and video belongs to the user.
- However, results made by AI may by chance resemble existing works. Please check for yourself before publication or commercial use. We do not warrant that outcome.
- A reference video is used to reference its visual grammar and narrative structure, and the service is designed not to copy the original’s story or scenes. However, it cannot prevent cases where a user enters input to follow the original as it is. Confirming rights in the reference video is the user’s responsibility.
4. What a person must check
AI can get facts wrong. Do not take numbers, dates, quotations or statements about people at face value; please check them before publication. That is why each step has a screen where you view and confirm the result.
5. Is it used for training
We do not use users’ inputs and results for model training. The data handling of outside AI services follows each company’s policy — you can look it up by the company names in the table above.
6. Cost
Generation steps trigger calls to outside services and incur costs. The screen shows the estimated usage before you start, and no call is made before you press the start button. In demo mode, saved example results are shown and no outside calls are made.
7. Contact
For inquiries about this policy, please write to official@aigorhythm.net.