Safety

Is OpenAI Dots Safe? Privacy & Controls (Unofficial)

What OpenAI says about dots safety: its own cloud computer, read-only background research, approvals, custom rules, and what you can pause, revoke or delete.

Unofficial guide · Last updated Sep 30, 2026Independent guide, not affiliated with OpenAI

Short answer

OpenAI says dots have built-in safeguards: each runs on its own cloud computer, your computer stays separate unless you connect it, background research uses read-only tools, and sensitive tasks like changing a password stay with you. You can set rules, pause or reset your dot. OpenAI also warns dots can still make mistakes.

We can't tell you dots is safe for your situation. This page lays out what OpenAI says the safeguards are, and where the risks are up to you.

What OpenAI says is built in

Separate computer. Each dot works on its own cloud computer. Your computer and its contents stay separate unless you choose to connect it.
Passwords. For signing in to supported websites, dots can use saved passwords without exposing them to the model.
Read-only background research. When you aren't working with it, your dot can look for ways to help using connected apps. Those tools are read-only, so it can't send messages, change app content or control your browser or computer.
Monitoring. OpenAI says security safeguards help defend against malicious instructions. If monitoring finds a concern, it can pause or stop the dot's work.
Approvals. Dots use auto-review to check actions that could affect your accounts or share information. Certain sensitive tasks, such as changing a password, always stay with you.

What you control

Apps. You choose which apps your dot can access and manage permissions in your ChatGPT app controls.
Custom rules. For each kind of action you can choose: take action without asking, take action if pre-approved, ask before acting, or hand off to you.
Computer access. Off by default. You turn it on with Allow access in the desktop app and can Revoke access at any time.
Pause and reset. Pause your dot from its profile. Reset deletes its conversations, saved memories and scheduled tasks.
Activity. Follow progress, including background work, in Activity View.

What to be careful about

Two things in OpenAI's own text are worth knowing before you connect anything.

First, your dot can review connected information and form memories from it even when you haven't asked a question. Second, disconnecting an app doesn't delete what it already got from that app. Only Reset does.

OpenAI also says dots can make mistakes, including when following your rules, so review consequential work. If you use it for anything with money or accounts, start with rules that ask first.

Texting

Texting uses a third-party provider and is a limited beta. OpenAI advises caution when sharing sensitive information through text messages.

Not sure yet if you can use dots? Check the eligibility checker →

FAQ

Can my dot read all my email?

Only if you connect it. You can connect your personal email so your dot can use it for tasks, and your dot can also review connected apps in the background to look for ways to help. Check each app's permissions before you connect it.

Are my dot's chats used to train OpenAI models?

On personal ChatGPT plans you can control whether your dot's conversations and work are used to improve models. OpenAI says it does not use Business, Enterprise or Edu workspace content that way by default.

Can my dot buy things without asking?

That depends on your rules. In Custom rules you choose whether your dot takes an action without asking, only if pre-approved, asks first, or hands it back to you.

What happens if I delete my dot?

Reset deletes its conversations, saved memories and scheduled tasks. Disconnecting an app does not delete what the dot already learned from it.

Sources

Last verified Sep 30, 2026

GetOpenAIDots is an independent, third-party guide and is not affiliated with, endorsed by, or sponsored by OpenAI. Facts on this page are checked against official sources where possible; unverified claims are marked accordingly.