Trust
Plain answers about AI, consent and your data.
The short version first. The legal version is linked at the bottom of every section.
AI labeling
Every assistant is AI, and every screen says so with a small AI label. Calynta never pretends to be a person, and assistants never claim to be your lawyer, accountant or doctor.
Consent and calls
Nothing goes to AI services until you agree, and your chats are private to you. Calls aren't recorded as audio or video. Calynta keeps a text transcript and a summary so it can pick up where you left off.
How we handle your data
Your data travels over encrypted connections. Passwords are stored only as secure hashes. Calls keep a text transcript and a summary. Calls are not recorded as audio or video. We don't sell your data, show ads, or train our own models on your conversations. Stored and processed in the United States. Delete your account in the app, or email hi@calynta.ai.
Model providers
Calynta uses more than one AI model behind the scenes. A new model can make your calls better without changing how they work. Providers: OpenAI, Anthropic, Google, Z.ai, Microsoft, OpenRouter, TypeSafe, LiveKit, AssemblyAI, Cartesia, Fish Audio, Groq, Anam. Also in the stack: Browserbase for the browser, RevenueCat plus Apple, Google, and Stripe for purchases. Nothing is sent until you agree in the app.
Deleting your data
Delete a single receipt, an assistant's memory of a call, or your whole account from settings. Deletion is permanent.
Questions about trust?
Write to us at hi@calynta.ai. A person will reply.