typecast
typecast · Audio and voice · KR · typecast.ai
services/typecast.yamlVendor data-handling terms
The Korean-law privacy policy sets out data categories, retention and cross-border transfers but does not say whether scripts or generated audio are used to improve the models.
“Destruction deadline: Personal information is destroyed without delay from the date it becomes unnecessary due to the expiry of the retention period, achievement of the processing purpose, discontinuation of the service, closure of business, and the like; where there is a legal obligation to retain it, it is stored securely for the relevant period. / Information related to facial images is not stored indefinitely and is retained only for the minimum period necessary to provide the service.”
The transfer table naming destination countries is rendered as a table on the page and is not quoted here.
“This Policy also applies to users who access the services from outside the Republic of Korea, including the European Economic Area (EEA), the United Kingdom, and the United States. / However, if you refuse a transfer that is essential to the provision of the service, your use of the service may be restricted.”
Domains and endpoints
observed means seen in Unseen deployments; vendor-documented means listed by the vendor. Vendors do not publish complete lists.
| Host | Role | Source |
|---|---|---|
| typecast.ai | app | observed |
Assessment
Reasoning: A Korean AI voice and avatar generator. The policy follows the Korean PIPA format with clear destruction rules and transfer notices, but does not address training on scripts or audio. Inputs are typically marketing copy or narration scripts.
Flags: Training-use of scripts and audio not stated · Cross-border transfers to unnamed countries in the quoted text · Assessed 2026-09-16. The assessment is Unseen's; the terms above are the vendor's.
Changelog
- 2026-09-16Documented. Retention and transfer terms verified against the Typecast privacy policy; training not stated.
- 2026-09-15Listed from the Unseen catalogue with observed domains.
Corrections
Pull request on GitHub, or the form below. Changes are reviewed and recorded in the changelog.