AI Data Policy

AI data handling

We do not use your data for model training. This policy defines data handling per product — Content Studio and API Platform.

GSCORA Content Studio data handling

• Prompts: used for generation; not stored in principle.
• Uploaded image / video / 3D files: used for generation and editing. Retention and deletion follow your account settings and this policy.
• Generated output: kept so you can use and save it; you can delete it from your account.
• Project data: stored for history and asset management.
• Model providers: prompts, uploads, and data needed for generation may be sent to and processed by the model providers you use.
• Training use: your data is not used for model training (provider policies differ per model).
• Overseas processing: data may be processed outside Japan depending on the model provider's region.

GSCORA API Platform data handling

• API requests (prompts & outputs): not stored in principle.
• Usage & cost metadata: stored on the control plane.
• Audit logs: metadata only — never prompts or generated content.
• BYOK credentials: stored encrypted with KMS.
• Model provider terms: follow each model's policy.
• Private Beta limits: data handling during Private Beta follows your individual agreement.

Per-model data policies

Processing region, retention, training terms, ZDR, and BYOK differ by model. See the Models page for details.

Subprocessors

For AI inference and generation, we may use the following model providers (subprocessors). The providers actually used depend on the models you select.

If we add or change subprocessors, we will update this page and notify in advance where required.

Processing countries & regions

The management infrastructure runs in a Japan region. Inference processing regions differ by model, as listed above.

Handling by data type

Operator: GSCORA Inc. (Representative: Representative Director: Zexin Yan, Address: 3-7-26 Ariake, Koto-ku, Tokyo 135-0063, Japan)