Tessera
Guide
Is it safe to tell AI personal information?
The honest version.
The useful answer is not yes or no. It depends on which terms you are under, who can read it later, and whether you can get it back out.
What happens to what you type
Your message goes to a model, an answer comes back, and a copy of the exchange is usually kept. The part that varies, and the part almost nobody checks, is what the company holding that copy is allowed to do with it.
Consumer terms and business terms are not the same
The free and personal tiers of most AI products reserve the right to use what you type to improve their systems. The business and API terms for the same models generally do not. Same model, same answer quality, different contract behind it.
This is the single biggest variable, and it is invisible from inside the chat window. If you want to know which one you are under, the answer is in the terms you accepted, not in the interface.
Three questions worth asking
1. Can you see what it has kept?
If there is no screen that shows you what a product remembers about you, you cannot audit it, and you are trusting a summary you have never read. Being able to open the memory and read every line is not a nice extra; it is the only way the other two questions can be answered honestly.
2. Can you correct it?
Memory that rewrites itself over time is convenient and slightly unnerving: the thing that describes you is being edited by something other than you. Ask whether you can change a specific fact, not just clear everything.
3. Can you get it out, and can you get rid of it?
Portability and deletion are the two exits. Without them, the sensible move is to share less.
A practical rule. Share what improves the answer. Withhold what would embarrass you in a data breach and would not change the reply. Most personal context that actually helps an AI is closer to "I have two kids and a commute" than to anything you would call sensitive.
How Tessera answers those three
Your memory opens like a settings page. You can read every line, change any of it, or delete it. Your questions reach the AI models on business tiers where conversations are not used to train the models by default, rather than the consumer versions where what you type can become training data. You can request a full copy of your data, delete a single fact, or delete the whole account, and deletion is real deletion.
Tessera is a private AI memory app. It is not an AI itself, which is why the question of what it keeps and what it sends is answerable at all: those are two separate things here, and you can look at both.
Questions
Related questions
Is it safe to tell AI personal information?
It depends on the terms you are under and whether you can see and remove what is kept. Consumer tiers of most AI products may use what you type to improve their systems; business terms for the same models generally do not.
Does AI train on what I tell it?
On consumer terms, often yes. Tessera never uses your conversations or your memory to train anything of our own, and your questions reach the AI models on business tiers where conversations are not used to train the models by default.
Can I delete personal information I have already given an AI?
With Tessera, yes: open your memory, find the fact, and delete it, or delete the whole account. With an AI app's own memory it varies, so it is worth checking before you rely on it.
Stop explaining yourself to AI.
Ten days free, no card required. iPhone and web.