Effective November 7, 2026
SwitchLingo is a product of Nova R&D Labs (novard.ai). The short version: no accounts, no ads, no selling your data. Your videos are deleted after 24 hours, and voice profiles are never stored.
When "Keep the speaker's voice" is on, SwitchLingo makes a temporary sound profile from the video's audio, used only to color that one voice-over. It stays in memory for that job, is never saved, and is never used to identify anyone. Recordings you contribute are kept as plain audio; we don't turn them into voiceprints or use them to recognize people. If you live in Illinois, Texas, Washington or another state with biometric privacy laws, don't use voice cloning or contribute recordings unless you agree to this use.
Hugging Face (hosting and the private dataset). Depending on the options you pick, text or audio may be sent to Microsoft (AI voices), Google (translation), Anthropic (Claude translation), ElevenLabs (premium voices; any cloned voice is deleted right after your job) or Replicate (lip-sync). They process data for us under their own terms.
Teen mode never clones voices, never accepts voice recordings, never uses premium third-party voices, and always hides swear words. The age group you pick is saved only in your own browser and attached to your translation job (deleted after 24 hours), so we can apply the right protections.
Videos we create carry a visible "AI translated · SwitchLingo" label, plus hidden metadata saying the file is AI-generated, by which system and when.
You can ask to see or delete what we hold about you, including your corrections and recordings, with the report form (choose "Privacy request"). California residents have rights under the CCPA and CalOPPA, and we honor them for everyone. We'll answer within 45 days.
We'll update this page if anything changes and change the date above.