- OpenAI Dots how to use: Dots are always-on AI assistants that keep working after you log off
- Core setup: Available to Pro and Business Premium users first; workspace admins enable Enterprise/Education access
- Key capability: Each Dot runs on its own cloud computer with a browser and over 4,000 plugins
- Security: You control app access with allow, approval-required, or block rules
- Current limit: One Dot per user for now; multi-Dot collaboration is a future roadmap item
What OpenAI Dots Are and How They Work
OpenAI Dots are always-on AI agents built into the ChatGPT ecosystem. Unlike a standard chat session, a Dot keeps working on your projects in the background even when your laptop is asleep, locked, or switched off. Each Dot runs on its own cloud computer with its own browser, powered by the GPT-6 Astra model under the hood.
Video Highlights:
- A Dot can manage several projects at once without separate chats
- Dots from teammates can work together toward a shared goal
- Dots have access to over 4,000 plugins
- A Dot can contact you for decisions when it needs one
- Powered by GPT-6 Astra with its own cloud computer and browser
The headline concept is simple: you delegate, the Dot executes. OpenAI positions Dots as assistants you do not have to babysit. You can check progress whenever you like, but repeated monitoring and correction are not required, which is the main departure from earlier agent tools.
Treat a Dot like a persistent teammate rather than a chatbot. Give it a project, walk away, and review results when it reports back instead of hovering over every response.
Core Capabilities Overview
Before diving into setup, it helps to understand what a Dot can actually do. The table below summarizes the main capability groups and how each one applies in practice.
| Capability | What It Does | Practical Example |
|---|---|---|
| Always-on work | Continues projects in the background | Prepares a report while your laptop is closed |
| Local file access | Interacts with files on your computer | Reads documents you authorize |
| Plugin access | Connects to 4,000+ plugins | Handles tasks across connected apps |
| Preference learning | Learns your goals and standards over time | Adapts output based on your feedback |
| Two-way contact | You message or call it; it reaches you too | Asks for approval before sending an invoice |
| Team collaboration | Teammate Dots can cooperate | Multiple Dots working toward one goal |
Personal Assistant
- Lifestyle support
- Notices forgotten tasks
- Prepares and sends invoices after approval
- Works across personal apps
Work Delegator
- Project ownership
- Runs multiple projects at once
- No separate chats to manage
- Reports progress unprompted
Team Player
- Cross-user collaboration
- Pairs with teammate Dots
- Shares goals and context
- Future of distributed work
Do not expect every response to be perfect. Dots learn from your feedback over time, so early corrections improve later results as the Dot internalizes your goals and preferences.
How to Use OpenAI Dots: Step-by-Step Setup
Follow these steps to get your first Dot up and running and establish a productive working routine with it.
Confirm Plan Eligibility
Check whether your account qualifies. Dots roll out to Pro and Business Premium users first. Enterprise, Education, and Healthcare users can try the beta if their workspace administrator enables the feature. Your first Dot is included with eligible plans.
Wait for Rollout
Availability is gradual, so even eligible accounts may not see Dots immediately. Watch your ChatGPT interface for the feature to appear after the announcement wave reaches your account.
Configure App Access Rules
Set permission rules for each connected app: allow outright, require manual approval, or block entirely. Define these boundaries before giving your Dot sensitive projects.
Assign Your First Project
Give the Dot a clear objective with context, preferred output standards, and any deadlines. One Dot handles multiple projects at once, so you do not need a separate chat per task.
Review the Activity View
Use the activity view to see exactly what your Dot is doing. Check for unauthorized app access attempts or unwanted actions, and course-correct through feedback so the Dot learns your preferences.
For now, each user can only have one Dot. Multi-Dot collaboration where several of your own Dots work together is on OpenAI's roadmap, but it is not available today.
Security and Control Settings
Security is a major focus of the Dots design, and OpenAI has layered several safeguards into the system. Because Dots are always on, transparency matters, especially for small and medium-sized businesses without large IT departments.
| Control Layer | Function | When It Triggers |
|---|---|---|
| App access rules | Allow, require approval, or block per app | Any time the Dot touches a connected app |
| Activity view | Live visibility into Dot actions | Continuous monitoring by the user |
| Review system | Flags sensitive or risky actions | When an action could compromise the account or expose sensitive data |
| Approval workflow | Human sign-off before execution | Actions the review system escalates |
A practical example from OpenAI: a Dot noticed a forgotten invoice, prepared it, and sent it only after the user approved it. That pattern, detect, prepare, request approval, execute, is the model to aim for with any sensitive workflow.
Never let a Dot send payments, invoices, or client communications without the approval-required rule enabled. The review system adds a safety net, but your own rules are the first line of defense.
Plan Eligibility and Getting Started Checklist
Not every ChatGPT user has access yet. This table breaks down who can try Dots and under what conditions.
| Plan | Access Status | Conditions |
|---|---|---|
| Pro | Rolling out | First Dot included with plan |
| Business Premium | Rolling out | First Dot included with plan |
| Enterprise | Beta available | Workspace admin must enable it |
| Education | Beta available | Workspace admin must enable it |
| Healthcare | Beta available | Workspace admin must enable it |
| Free / Plus | Not yet announced | No confirmed timeline |
Getting Started Checklist:
- Verify your plan is Pro, Business Premium, or an admin-enabled workspace
- Wait for the Dots feature to appear in your ChatGPT account
- Set app access rules before assigning sensitive projects
- Assign your first project with clear goals and standards
- Review the activity view daily during the first week
- Give feedback so the Dot learns your preferences
Start with low-risk tasks like research summaries or drafting documents. Gradually expand to higher-stakes workflows only after the Dot's output consistently matches your standards.
FAQ
Q: How do I use OpenAI Dots if I have a Pro account?
Pro users get access as the feature rolls out gradually. Your first Dot is included with the plan, along with an allowance for deeper work. Once Dots appear in your ChatGPT interface, you can assign projects immediately.
Q: Do Dots keep working when my computer is off?
Yes. Each Dot runs on its own cloud computer with its own browser, so it continues working whether your laptop is asleep, locked, or shut down. You can check progress at any time, but constant monitoring is not required.
Q: Can I have multiple Dots?
Currently you can only have one Dot. OpenAI has stated that access to multiple Dots working together is a future direction, and demos have already shown teammate Dots collaborating across users.
Q: How do I stop my Dot from accessing an app?
Open the control settings and set the app's rule to block, or set it to require approval so you sign off each time. The activity view also shows if a Dot attempts access you never authorized.