What an AI Assistant Should and Should Not Read
An assistant that prepares work needs context. Here is what it should read, what it should not, and how admins set the limits.
Why does an assistant need to read anything?
To prepare useful work, it needs the same context a person would use: the email thread, the calendar invite, the chat message. Without it, the draft is generic.
What does Jin read?
The email, calendar, chat, and files your team connects, used only to prepare work for that person. Each employee connects their own account and can disconnect at any time.
What does Jin never do?
- It never sends or changes anything without a person's approval.
- It never records your screen.
- It does not score or rank people, and what it finds goes to the employee, not their manager.
Want to see it on your team's work?
- One team, one task to start
- No prompts to learn
- Your data is never used to train AI models
How do admins set the limits?
Your admin chooses which teams and tools are connected, and can exclude any folder, channel, label, or team before anyone starts.
How is company data protected?
Customer data is not used to train AI models, and each company's data is kept separate. When improvements are shared across a team, the way of doing the task is shared, not anyone's emails or files, and an admin reviews it first.
See the security page on jinai.md for details.
FAQ
Is our data used to train AI models?
No. Customer data is not used to train AI models, and each company's data is kept separate.
Can employees turn it off?
Yes. Each employee can disconnect their account at any time.