Klint.ai

Klint insurance guide

Privacy and Data Handling

Short answer: Policy text sent for an AI explanation is processed by the AI service. Original upload files are not kept as policy attachments. Optional accounts store the details, selected evidence and plans you choose to save.

Last updated . Klint educational content; not licensed advice.

What is processed during a conversation

Questions and supplied text are sent to the application and AI service to answer your request. Avoid unnecessary personal identifiers, medical details, passwords or financial account information. Client-side redaction is a convenience, not a guarantee that all sensitive information has been removed.

App policy storage is separate from AI-provider handling. Provider and platform retention terms need legal review; Klint does not promise zero vendor retention or confidentiality beyond verified terms.

What an optional account can store

Saved policy metadata, entered notes, explicitly selected evidence, follows, reminder preferences, action plans and minimal activity records can be stored in the account database. Raw file bytes are not saved as policy attachments. An account is not needed for a generic question.

Account authentication uses Clerk browser/session storage. The optional measurement choice is remembered on your device. These preferences are distinct from your insurance records; provider storage and retention terms still require legal review.

Export and deletion

My insurance includes an export of your own saved data and a typed-confirmation account deletion flow. Deletion removes owned private records and the account identity; a downloaded export remains on your device. Published sourced public records may remain without reviewer attribution.

If identity-provider deletion fails after private data has been removed, the app reports that partial failure instead of claiming nothing was deleted. Active-record deletion is not a promise of immediate removal from every service log, backup or external processor; those retention terms require legal review.

Offline and usage measurement

The PWA caches public static assets and generic connection-required help, not policy files, private API responses, chats or personalized insurance answers. Answers and saved records need a connection.

First-party daily usage counts are off by default and require the optional measurement choice. They contain only an allowlisted event, public page, coarse referrer class and count—not question text, filenames, quotes, health notes, account identifiers or full referral URLs. Browser Do Not Track and Global Privacy Control keep measurement off. No insurer lead tracking is enabled. Aggregate counts are retained for reporting; an automated retention purge is not configured. Network metadata may be processed for ordinary hosting and security. Consent and retention requirements still need legal review.

Ask Klint a question about your own policy