ChatGPT Dot (2026): What It Is, Who Gets One and How to Set It Up

So if you're on Pro in the UK or the EU, you won't see one yet. For everyone else, here's what a dot does, how to set it up, and the settings I'd look at before trusting it with anything that matters, all from OpenAI's own Help Center.

What a dot does that a normal chat doesn't

A regular chat waits for you. A dot takes on ongoing responsibility. You give it a goal and the context, and it keeps making progress between conversations, then brings results back for your review. It works across the apps you choose to connect, starts with context from your ChatGPT memory and learns how you like to work. It turns to you for decisions that need your judgement.

OpenAI's own examples are business-friendly: pull together data and customer feedback and update a presentation as new information comes in, turn an interview transcript into show notes and social posts and carry your edits across the drafts, or scope, code and test a change to an app. A dot can also create cloud tasks in Codex cloud environments, as long as you've created the environment in Codex first. If the line between Chat, Work and Codex is still fuzzy, sort that out first.

Who can get a dot as of Nov 2026

Plan	Availability
Pro	Users aged 18 and older, outside the European Economic Area, Switzerland and the UK. First dot included at no extra cost.
Business Premium	Supported ChatGPT regions, subject to workspace eligibility. First dot included at no extra cost.
Enterprise, including Edu and Healthcare	Beta only, and off by default until the workspace admin turns it on.

Your plan also includes an allowance for deeper work, with extended limits for the first month after launch. OpenAI says you'll later be able to add more dots, or scale one by increasing its speed or the amount of work it can take on each month. Rollout is gradual and may take several days to reach your account. If you can't find it, check you're on the intended account and workspace, then open chatgpt.com/dots. For the plans themselves, see how the ChatGPT Pro tiers compare.

Setting up your dot

1. Open the ChatGPT mobile app on iOS or Android, the desktop app (which is also on Windows) or ChatGPT on desktop web. Mobile web isn't supported.
2. Follow the onboarding prompts.
3. Give it a name if you like. Its default handle is @yourname-dot, and naming it changes that to @yourname-agentname.
4. Open its profile in the desktop app to see work under In progress, Scheduled and Completed, and to open its computer to watch or step in.
5. Set your custom rules before you hand it anything that sends or spends (more on that below).

Apps, memory and your own computer

Connected apps give a dot information for your tasks, and the help page flags something easy to miss: a dot can review that information proactively and form memories from it, even when you haven't asked about it. Disconnecting an app doesn't delete what it's already picked up. The only way to delete a dot's own saved memories is to delete the dot. It also receives memories from ChatGPT, so it's worth knowing how ChatGPT Memory works before you start.

You can connect Slack from ChatGPT on desktop, and connect your personal email so the dot can use it for your tasks. At launch it can't have its own standalone email address, and it can't call you.

Access to your local computer is optional and starts turned off. To grant it, use the ChatGPT desktop app on that computer and confirm Allow access. Confirm Revoke access to stop it. With access on, the dot can work on files there, create Work or Codex tasks, use local skills and use your local browser when its cloud browser is blocked. Keep that computer online with the ChatGPT app open.

One quirk to know: the dot's cloud browser has its own sign-in sessions. Signing in on your own computer doesn't sign the dot in. If a website blocks the cloud browser, you can ask the dot to try your connected computer instead.

Recurring checks and reminders

You can ask a dot to set a reminder or run a recurring check, such as reviewing your calendar each morning and telling you what's coming up. Its scheduled work sits under Scheduled in its profile. If you've used scheduled ChatGPT tasks, the idea is familiar; the difference is that your dot does the checking, with whatever apps you've connected.

Texting your dot and the encryption catch

Texting is a beta for a limited number of Pro users in the US. It isn't supported in Business or Enterprise workspaces, and you need iMessage or RCS. To turn it on:

1. Open your dot's profile and select Texting.
2. Enter your phone number and select Send confirmation text.
3. Enter the code you receive and select Verify phone number.

Now the catch. OpenAI uses Linq, a third-party messaging provider, to send and receive these texts. Because that provider processes the messages on the way to OpenAI, the conversation isn't end-to-end encrypted, even over iMessage or RCS. The provider doesn't keep message contents after delivery, but OpenAI stores your texts as part of the dot's conversations. I wouldn't text it anything I'd hesitate to put in an ordinary email.

Other limits: it's one-to-one only with no group texts, SMS isn't supported, and you can't call the dot's number. Calls happen through the phone button in the ChatGPT mobile app. Reply STOP to stop texts. Turning texting off disconnects your number but doesn't delete past conversations or memories.

Custom rules decide what it does alone

On mobile, go to Customize, then Custom rules, to see the defaults or add your own. Each rule gets one of four behaviours:

• Take action without asking.
• Take action if pre-approved, where pre-approved means you explicitly asked for that action in your prompt.
• Ask before taking action.
• Hand off to you.

Three lines on the help page are worth pinning above your desk. Asking for a draft doesn't authorise sending it. Custom rules don't grant app or computer access or override built-in safeguards. And some steps, such as changing a password, must be done by you. OpenAI also says plainly that a dot can make mistakes, including when following your rules, so review its work and check important details before relying on it.

Pausing or deleting a dot

To pause, open the more options menu in its profile and select Pause. To restart, select Paused, then Tap to resume.

Deleting is more final than it sounds. It removes the dot's own saved memories and scheduled tasks and archives your main conversation, so you can't continue it with that dot. Any separate Work or Codex chats it created are kept. Deletion can't be undone, and it doesn't create a replacement.

What I'd do first

I'd start with one low-risk recurring job, like the morning calendar check, and see how it reports back. I'd keep Ask before taking action on anything that sends, posts or spends. And I'd skip local computer access until I trust how it behaves on its own cloud computer, since you can switch that on later. Vendor features change often, so check OpenAI's own help page for the latest on availability.

If you want an AI agent doing real work for your team with clear guardrails, that's the kind of AI enablement I help with. I run AI training and workshops for teams, and help businesses with AI enablement and workflow automation, so the busywork runs itself and people get time back. Most of that work is with SaaS marketing teams and small businesses. Here's how we can work together.

If you want the full set, start at the ChatGPT Prompts and Limits: 30 Guides to Fun Prompts, Photo Edits and Daily Caps.

Frequently asked questions

Is ChatGPT dot available in the UK?

Not for Pro users as of Nov 2026. OpenAI's help page says Pro dots are rolling out outside the European Economic Area, Switzerland and the UK.

Does a dot cost extra?

Your first dot is included in Pro or Business Premium at no extra cost, according to OpenAI's help page. OpenAI says you'll later be able to add more dots or scale each one.

Are texts to my dot encrypted?

Not end-to-end. A third-party messaging provider processes the texts on the way to OpenAI, so they aren't end-to-end encrypted even over iMessage or RCS.

Can a dot use my own computer?

Only if you allow it. Local computer access starts turned off, and you grant it through the ChatGPT desktop app on that computer with Allow access, or stop it with Revoke access.