Let's address the elephant in the room.
You're considering letting an AI analyze your text messages. Your actual conversations. With real people. And somewhere in the back of your mind, a very reasonable voice is asking: What happens to all those phone numbers?
It's a fair question. Your contact list is one of the most personal things on your phone — arguably more sensitive than your photos. Every number represents a real relationship. A parent. A best friend. A therapist. Someone you're dating.
So here's the honest answer about how Amicai handles that data.
What We Actually Need (And What We Don't)
Amicai analyzes the content of your conversations — the topics you discuss, the emotional patterns, the life events you mention, the frequency of contact. That's what powers your relationship insights.
What we don't need — and actively work to keep out of sight — is the raw identifier data that connects those conversations to specific phone numbers.
Think of it this way: your daily reflection might say "You and Devon had a meaningful conversation about his recovery." It will never say "You and +1 (860) 303-7149 had a meaningful conversation."
Three Layers of Protection
We don't rely on a single security measure. Phone number protection works at three separate levels:
At the data layer, when your messages are processed, phone numbers are used internally to match conversations to the right contact names — then stripped from any output that reaches you. The AI that generates your insights never sees raw phone numbers in its input.
At the application layer, every screen in Amicai filters content before rendering. Participant lists, conversation threads, event details — all of it passes through checks that catch and remove phone-number patterns before they appear on screen. Even if something unexpected slipped through the data layer, the application layer catches it.
At the AI layer, the language model that powers your chat assistant and daily reflections receives contact names, never phone numbers. It can tell you about your relationship with "Devon" without ever knowing Devon's number.
Why This Matters More Than You Think
Most apps that handle contact data don't think about phone number exposure because — in most contexts — seeing a phone number isn't a big deal.
But Amicai is different. Your relationship insights live on a screen. Maybe your partner glances at your dashboard. Maybe you show a friend your weekly reflection. Maybe you share a screenshot of a particularly thoughtful AI insight.
In any of those moments, a visible phone number becomes a privacy leak. Not just yours — someone else's. Someone who never signed up for Amicai and never consented to having their number displayed on your screen.
That's the standard we hold ourselves to: the people in your contact list deserve the same privacy consideration you do, even though they're not our users.
What We Don't Store
A few things Amicai deliberately does not keep:
-
No images or media. We process text only. Photos, videos, and attachments are never stored. If you shared an image in a conversation, our system shows a placeholder: "[Media shared — content not stored for privacy]."
-
No message content in our databases. Your raw messages are processed for insights and then the analysis is stored — not the messages themselves.
-
No contact list exports. Your contacts exist in Amicai to power your experience. They're never shared, sold, or used for advertising.
The Honest Caveat
No system is perfect. We're a small team building something new, and security is an ongoing practice, not a finished product.
What we can tell you is that phone number protection isn't an afterthought bolted on at the end. It's built into every layer — from how we process data, to how our AI generates responses, to how our frontend renders content. When we find gaps, we fix them immediately and strengthen the surrounding systems.
We also believe in transparency. For a deeper look at how commercial AI APIs handle data differently from consumer chatbots, see what AI companies actually do with your data. If you ever see something that looks like a phone number somewhere it shouldn't be, we want to know about it. That kind of feedback makes the product better for everyone.
The Bottom Line
Your relationships are deeply personal. The data that represents them should be handled with care — not just your data, but the data of every person in your life who shows up in your conversations.
That's the commitment. Not a vague "we take privacy seriously" statement, but a specific, multi-layered system designed to keep phone numbers out of places they don't belong.
Your conversations deserve real analysis. The people in those conversations deserve real privacy. We think you can have both. Here's what that analysis actually looks like when someone tries it.