Privacy Policy
1. User Information Collected and How It Is Collected
(1) Information provided by users
- • Email address
- • Information about payment methods, such as credit card information, bank account information, and electronic money information
- • Still image information that includes the user's likeness
- • Information that users enter or send through input forms or other methods specified by the Operator
(2) Information provided from other services when users allow the Service to link with those services
- • The ID the user uses on that external service
- • Other information that the user has allowed to be disclosed to linked services through the privacy settings of that external service
(3) Information the Operator collects when users use the Service
- • Referrers
- • IP addresses
- • Information about server access logs
- • Cookies, ADID, IDFA, and other identifiers
(4) Information the Operator collects when users allow integration with external AI clients (ChatGPT, Claude, etc.)
- • The name, version, and User-Agent of the connecting AI client
- • The names of the features (tools) used and the identifiers of the works and episodes concerned
- • The amount of data sent and received, processing time, and error information
- • IP addresses
2. Purposes of Use
- (1) To provide, maintain, protect, and improve the Service, including accepting registrations, verifying identity, authenticating users, recording user settings, and calculating fees for payment
- (2) To measure user traffic and behavior
- (3) To deliver and display advertisements and measure their effectiveness
- (4) To provide information about the Service and respond to inquiries
- (5) To respond to acts that violate the Operator's terms, policies, and the like relating to the Service (the "Terms, etc.")
- (6) To notify users of changes to the Terms, etc. relating to the Service
- (7) To send email notifications about interactions between users within the Service, such as follows, likes, reviews, updates to followed works, and new works being published
- (8) To analyze data needed to improve the Service or media operations
- (9) To analyze publicly available content (including posts such as reviews and comments), including with AI, for work recommendations, curation, suggesting tags and genres, ensuring the safety of public content including monitoring for defamation, and providing authors with information based on statistical analysis
- (10) To provide the integration feature with external AI clients (ChatGPT, Claude, etc.) that the user has allowed to connect
- (11) To calculate the per-work AI usage classification (an automatic classification based on the record of AI feature use) and display it as information about the work
- (12) To analyze novel text, including private text, with AI to automatically identify the speaker of each line of dialogue, and for the Operator to evaluate and improve the accuracy of that identification
- (13) To check how AI was used in works submitted to contests and similar events
3. Methods of Notification, Publication, or Obtaining Consent, and How to Request Suspension of Use
4. External Transmission
About training AI models
All AI-related services used by the AI features of the Service are used through APIs, and the information sent is not used to train AI models. In addition, for some services (such as Voyage AI), the Operator has also opted out so that the information is not used to train AI models.
However, this does not apply when users themselves connect an external AI client (ChatGPT, Claude, etc.) to the Service. How the information is handled by the connected service (including whether it is used to train AI models) follows the settings and privacy policy of that service. For details, see "5. Integration with External AI Clients (MCP)".
Information is sent to AI features only when users explicitly use an AI feature, for features the user has agreed to in advance, and for the automatic identification of dialogue speakers described next.
For the automatic identification of dialogue speakers, episode text (including unpublished drafts) and character names and nicknames are automatically sent in the background, without any user action, to the external services listed under "(1) AI features" below, which identify the speaker of each line of dialogue. The Operator may use the identification results and the text to evaluate and improve the accuracy of the identification. Works whose AI feature setting is "Do not use AI" are not sent.
In addition, publicly available content (including posts such as reviews and comments) is subject to analysis, including with AI, for purposes such as work recommendations, curation, suggesting tags and genres, ensuring the safety of public content including monitoring for defamation, and statistical analysis. Private content (data in the editor) is not subject to analysis, except when the user explicitly uses an AI feature, when tag and genre candidates are suggested on the screen for publishing a work, and for the automatic identification of dialogue speakers described above.
(1) AI features
When AI features such as AI writing assistance and semantic search are used, and when the speakers of dialogue are automatically identified, information is sent to the following services.
| Service | Provider | Purpose | Information sent |
|---|---|---|---|
| OpenAI API | OpenAI, LLC (United States) | Image generation and providing AI writing assistance features | Prompts and images |
| Anthropic API | Anthropic PBC (United States) | Providing AI writing assistance features | Input text, prompts, and images |
| Voyage AI | Voyage AI, Inc. (United States) | Providing semantic search features | Novel text and search queries |
| Google Vertex AI | Google LLC (United States) | Image generation and providing AI writing assistance features | Input text, prompts, and images |
| xAI (Grok) | xAI Corp (United States) | Providing AI writing assistance features | Input text, prompts, and images |
| TypeSafe API (Jev) | TypeSafe AI, Inc. (United States) | Automatic identification of dialogue speakers, suggesting tags and genres for works, and checking the safety of posted content | Novel text (including unpublished drafts), character names and nicknames, and posts such as reviews and comments |
| Cloudflare Workers AI | Cloudflare, Inc. (United States) | Automatic identification of dialogue speakers, suggesting tags and genres for works, and checking the safety of posted content | Novel text (including unpublished drafts), character names and nicknames, and posts such as reviews and comments |
| Vercel AI Gateway | Vercel Inc. (United States) | Automatic identification of dialogue speakers, suggesting tags and genres for works, and checking the safety of posted content | Novel text (including unpublished drafts), character names and nicknames, and posts such as reviews and comments |
(2) Analysis and improvement
The following services are used to analyze how the Service is used and to improve its quality.
| Service | Provider | Purpose | Information sent |
|---|---|---|---|
| Google Tag Manager | Google LLC (United States) | Analyzing and improving how the Service is used | Page views and event information |
| Sentry | Functional Software, Inc. (United States) | Error monitoring and improving service quality | Error information and session information |
(3) Payment
The following service is used to process payments for paid services.
| Service | Provider | Purpose | Information sent |
|---|---|---|---|
| Stripe | Stripe, Inc. (United States) | Payment processing | Name, email address, and payment information |
(4) Community
The following service is used for community features.
| Service | Provider | Purpose | Information sent |
|---|---|---|---|
| Discourse | Civilized Discourse Construction Kit, Inc. (United States) | Community integration | User ID and email address |
(5) Infrastructure and data storage
The Service's operating infrastructure (hosting, data storage, aggregation of application logs, etc.) uses cloud services located in the United States. As contractors of the Operator, these providers handle information only within the scope of the Operator's instructions and do not use its content for their own purposes. The countries where the information is handled are as follows.
| Category | Country where handled | Information handled |
|---|---|---|
| Hosting and content delivery | United States | Access logs, IP addresses, and request information |
| Storage of databases and files | United States | All information stored in the Service (including uploaded images) |
| Access control and storage of authentication information | United States | IP addresses and authentication information |
| Aggregating application logs, investigating failures, and sending operational alerts | United States | Access logs, IP addresses, request information, and error information |
| Platform for analyzing usage data | United States | Activity history and hashed identifiers |
In addition to the above, the Service uses the following services, whose providers handle user information itself. These providers also handle information only within the scope of the Operator's instructions, as contractors of the Operator.
| Service | Provider | Purpose | Information sent |
|---|---|---|---|
| Firebase Authentication | Google LLC (United States) | User authentication platform | Email addresses and authentication information |
| Resend | Resend, Inc. (United States) | Email delivery | Email addresses and notification content |
5. Integration with External AI Clients (MCP)
(1) Information sent and where it is sent
When the user performs an operation on an AI client they have allowed to connect, the following information is sent to the connected service in response to that request. If the user has not allowed a connection, or has not performed any operation on the AI client after connecting, no information is sent.
| Service | Provider | Purpose | Information sent |
|---|---|---|---|
| ChatGPT | OpenAI, LLC (United States) | Writing assistance from AI clients that the user has allowed to connect | Work information (title, synopsis, AI feature settings, etc.), chapter (section) plots, episode text (including private drafts), characters and terms (including attribute values and nicknames), documents, foreshadowing, relationship diagrams, and account information (user ID, display name, and avatar image URL) |
| Claude | Anthropic PBC (United States) | Writing assistance from AI clients that the user has allowed to connect | Work information (title, synopsis, AI feature settings, etc.), chapter (section) plots, episode text (including private drafts), characters and terms (including attribute values and nicknames), documents, foreshadowing, relationship diagrams, and account information (user ID, display name, and avatar image URL), and maps (including images) |
| Other AI clients | The provider of each client | Writing assistance from AI clients that the user has allowed to connect | Work information (title, synopsis, AI feature settings, etc.), chapter (section) plots, episode text (including private drafts), characters and terms (including attribute values and nicknames), documents, foreshadowing, relationship diagrams, and account information (user ID, display name, and avatar image URL), and maps (including images) |
This integration is based on the standard specification of MCP (Model Context Protocol), and AI clients other than those listed above can also connect if they support it. Information is sent to the provider of the AI client the user has allowed to connect.
The work data sent includes drafts that have not yet been published. On the other hand, email addresses, payment information, and authentication information such as passwords are not sent.
(2) Operations the connected AI client can perform
In addition to reading the above information, the connected AI client can, based on the user's instructions, create and update works, and create, update, and delete episodes, characters, terms, documents, foreshadowing, relationship diagrams, and chapters (sections). These operations remain possible until the user disconnects the integration.
(3) Handling by the recipient
The storage and use of the information sent (including whether it is used to train AI models) follow the user's settings on the connected service and that provider's privacy policy. The Operator does not control how the information is handled by the connected service. Please check the privacy policy of the connected service before allowing a connection.
The privacy policies of the main AI clients are as follows (as of the revision of this Policy; please check each provider's website for the latest version).
- ChatGPT (OpenAI, LLC): OpenAI Privacy Policy
- Claude (Anthropic PBC): Anthropic Privacy Policy
When connecting any other AI client, please check that provider's privacy policy.
(4) Disconnecting the integration
The integration can be disconnected at any time from the settings screen of the connected AI client. After disconnection, issued authentication information expires within 30 days, after which the connected service can no longer access data in the Service. If the disconnection is not reported to the Operator by the connected service, the authentication information remains valid until the expiration period above has passed.
(5) Information the Operator collects and records during integration
To provide and improve the integration feature and to prevent misuse, the Operator records the name and version of the connecting AI client, the names of the features used, the identifiers of the works and episodes concerned, the amount of data sent and received, processing time, error information, and IP addresses. For analysis purposes, the user identifier used in these records is not stored as the user ID itself, but is converted into a hashed identifier before being stored.
In addition, for tools that create or update episodes, the episode text sent is stored as evidence for the per-work AI usage classification. Both the text and the records are deleted when the user withdraws or deletes the data.
6. Retention Periods
| Information | Retention period |
|---|---|
| Account information and work data (text, characters, terms, etc.) | Until the user deletes their account or the data |
| Records of AI feature use (records of the date and time of use, the work concerned, and the type of feature, which form the basis of the per-work AI usage classification) | Until the user deletes their account or the data |
| Records of episode text sent through the integration feature (MCP) (which form the basis of the AI usage classification) | Until the user deletes their account or the data |
| Authentication information for external AI client integrations (access tokens, etc.) | Expires within 30 days after the integration is disconnected |
| Usage statistics of external AI client integrations (tool names used, data volume, processing time, etc.) | 400 days |
| Data for analyzing how the Service is used (history of page views and operations, referrers, identifiers made by hashing IP addresses, etc.) | 400 days |
| Application logs | 30 days |
| Error information (Sentry) | 90 days |
7. Provision to Third Parties
- (1) When the Operator entrusts all or part of the handling of personal information to another party within the scope necessary to achieve the purposes of use
- (2) When personal information is provided as a result of the succession of business due to a merger or other reasons
- (3) When it is necessary to cooperate with a national government agency, a local government, or a party entrusted by either of them in carrying out affairs prescribed by laws and regulations, and obtaining the user's consent may interfere with carrying out those affairs
- (4) Other cases permitted by the Act on the Protection of Personal Information (the "Personal Information Protection Act") or other laws and regulations
8. Disclosure of Personal Information
9. Correction and Suspension of Use of Personal Information
- When a user requests, under the Personal Information Protection Act, (1) correction of their personal information on the grounds that it is untrue, or (2) suspension of its use on the grounds that it is being handled beyond the scope of the purposes of use published in advance or that it was collected by deception or other wrongful means, the Operator will, after confirming that the request comes from the user themselves, conduct the necessary investigation without delay, correct the personal information or suspend its use based on the results, and notify the user accordingly. If the Operator decides not to correct it or suspend its use, the Operator will notify the user of that decision.
- When a user requests deletion of their personal information and the Operator determines that it needs to comply with the request, the Operator will, after confirming that the request comes from the user themselves, delete the personal information and notify the user accordingly.
- Sections 9-1 and 9-2 do not apply where the Operator has no obligation under the Personal Information Protection Act or other laws and regulations to carry out the relevant correction, suspension of use, or other such measures.