AI Companion Privacy: What to Check Before You Chat

Start with the provider's own privacy policy. Look for categories such as account details, conversation content, uploaded media, voice data, device information, usage activity, purchases, and support messages. The exact categories and purposes vary by service.

Memory features may save profile facts or summaries separately from the current conversation. Check whether the product offers controls to view, edit, or delete remembered information and whether those actions affect other stored records.

A privacy review should ask why data is processed, whether service providers or other parties receive it, how long it is retained, and what choices the user has. Do not assume that a private-feeling conversation is confidential in the legal or professional sense.

Find the account deletion and privacy-request process before relying on the service. Removing a message from the interface, deleting a character, and deleting an account may be different actions with different retention effects.

Use a unique password and available account security features. Avoid sharing credentials, financial information, precise location, private records, or another person's identifying information. Consider using less specific biographical details when exact details are not needed for the intended experience.

Review permissions for microphone, camera, photos, notifications, and contacts. Grant only access required for features you choose to use, and revisit permissions when the app adds features.

NIST frames privacy risk management as an ongoing organizational process. For an individual user, the practical equivalent is to recheck policies, settings, and retained information periodically rather than treating initial consent as permanent understanding.

CompanionIndex records verification dates because product policies change. Always use the provider's current policy as the controlling source and seek qualified advice if a specific legal or regulatory question matters.