Threat models matter.
When a platform/service/app tells you they are “private” or “secure” always ask “from whom?”
Criminals, domestic abusers, law enforcement, data brokers, and intelligence agencies are all different attackers with very different capabilities.