# ChatGPT Dots for Business: What They Are, Who Gets Them, and 5 Office Tasks to Hand Off

> ChatGPT dots are always-on agents that keep working between chats. Who can get them, what they cost, 5 office tasks to hand off and the guardrails to set.

Source: https://learn.ynteractive.com/content/explainers/chatgpt-dots-for-business · Updated 2026-10-14 · Free from Stepthrough (https://learn.ynteractive.com)

By [Robert Breen](https://learn.ynteractive.com/content/about) · Published October 14, 2026 · Updated October 14, 2026 · 11 min read

A [ChatGPT dot](https://learn.ynteractive.com/content/glossary/chatgpt-dot) is an [always-on agent](https://learn.ynteractive.com/content/glossary/always-on-agent) that keeps working on a goal between your conversations. [OpenAI](https://learn.ynteractive.com/content/glossary/openai) launched dots on September 29, 2026. Each dot runs on GPT-6 Astra with its own cloud computer, can connect to over 4,000 apps through plugins, and can be reached in ChatGPT, Slack and Microsoft Teams. Right now they are only on Pro, Business Premium and Enterprise (as an admin-enabled beta), and Pro access excludes the EEA, Switzerland and the UK. If you are on Plus, you can get much of the same value today with [ChatGPT Work and a scheduled task](https://learn.ynteractive.com/content/chatgpt-work-multi-step-task), which this page covers at the end.

This page is for office managers, owners and team leads deciding whether dots are worth paying for, and what to do with one. Every feature and plan detail below was checked against OpenAI's own announcement and help center on October 14, 2026. We have not run a dot ourselves (our accounts are on Plus), so we describe only what OpenAI documents.

## What business problem do dots solve?

Most office work is not hard. It is recurring. Someone has to remember to chase the vendor who has not confirmed a delivery, check the invoices inbox, look at what competitors changed this week and update the client deck when the scope moves. Each task takes minutes, but they pile up, and the cost shows up as the follow-up nobody sent and the invoice that went out three weeks late.

A normal AI chat does not help much with that, because it waits for you to ask. A dot is built for the opposite: you give it an ongoing responsibility, and it keeps track, checks for changes and comes back to you when a decision needs a person. OpenAI's own example is an early tester whose dot noticed he had forgotten to invoice a client, prepared the invoice and sent it after he approved it. That is the business case in one line: fewer things fall through the cracks, and a person still signs off.

## What exactly is a ChatGPT dot?

OpenAI calls dots "always-on agents built to handle everything." In plain terms, a dot is an [AI agent](https://learn.ynteractive.com/content/glossary/ai-agent) that you keep, rather than one you start for a single task. The details OpenAI documents:

- **Its own computer.** Each dot works on its own cloud computer with its own browser. You can open that computer from the dot's profile to watch or take over. Your own laptop stays separate unless you choose to connect it, and local access starts turned off.
- **Your apps.** Through plugins it can connect to over 4,000 apps. You choose which ones, and permissions follow your existing ChatGPT app controls.
- **Memory.** It starts with context from [ChatGPT memory](https://learn.ynteractive.com/content/glossary/chatgpt-memory), creates its own memories (including from connected apps) and learns how you like work done from your feedback.
- **Where you reach it.** ChatGPT on desktop, web and mobile, plus Slack and Teams. You can also call it by voice. Texting is a limited beta for some US Pro users.
- **Background work.** When you are not working with it, a dot does what OpenAI calls proactive research, using your connected apps with read-only tools that cannot send messages, change app content or control your browser or computer.
- **Recurring work.** You can ask it to run a reminder or a recurring check, and you review its work under In progress, Scheduled and Completed in its profile.

You start with one primary dot, give it a name and pick a character or pet for it. OpenAI says you will be able to add more dots later.

## How is a dot different from a chat, agent mode, Projects and scheduled tasks?

The short version: a chat answers, [ChatGPT Work](https://learn.ynteractive.com/content/chatgpt-work-multi-step-task) (which replaced the older agent mode in July 2026) finishes one job, a Project remembers your files and instructions, a scheduled task repeats one job on a timer, and a dot owns an ongoing goal and decides what to do next.

| Tool | What it does | Who starts the work | Keeps going between chats? |
| --- | --- | --- | --- |
| Normal chat | Answers a question or drafts something | You, every time | No |
| ChatGPT Work (agent mode) | Finishes one multi-step job, such as a spreadsheet from three websites | You | Only if you schedule it |
| ChatGPT Projects | Keeps files, instructions and chats on one topic together | You | It remembers, but does not act |
| Scheduled task | Runs the same saved job at a set time | The timer | Yes, the same job each time |
| ChatGPT dot | Takes an ongoing goal, watches for changes, plans next steps, asks for approval | The dot, within your rules | Yes |

The [comparison of dots, scheduled tasks, agent mode and the Claude and Gemini equivalents](https://learn.ynteractive.com/content/compare/chatgpt-dots-vs-scheduled-tasks-vs-agent-mode) goes deeper, and the free lesson [AI Agent vs Chatbot vs Automation](https://learn.ynteractive.com/content/ai-agent-vs-chatbot-vs-automation) shows the difference with one real job.

## Who can get dots, and what do they cost?

As of October 14, 2026, OpenAI's help center lists:

- **Pro:** users over 18, outside the European Economic Area, Switzerland and the UK. OpenAI's pricing page lists Pro from $100 a month, in three usage tiers ($100, $200 and $500).
- **Business Premium:** across all supported ChatGPT regions. OpenAI says Premium seats include 5 times more usage, no five-hour limit and access to dots. OpenAI's business page does not show the seat price; third-party pricing guides report about $100 per seat per month on annual billing, so confirm it with OpenAI before you budget.
- **Enterprise, Edu and Healthcare:** a beta that a workspace admin must turn on. It is off by default.
- **Not included:** Free, Go and Plus. OpenAI says it plans to expand dots to more users but has not given a date.

Your first dot is included in a Pro or Business Premium plan at no extra cost. Talking to your dot does not count toward your ChatGPT usage limits. Deeper work comes out of an allowance in your plan, with extended limits for the first month after launch, and tasks your dot starts in ChatGPT Work or Codex count toward those products' limits as usual. OpenAI has not published the size of that allowance. Rollout is gradual, so access can take several days to reach an account; the help center says to check chatgpt.com/dots.

A note on conflicting reports: some coverage says dots run on "GPT-6." OpenAI's own announcement and help center say GPT-6 Astra, so that is what we use here.

## What are 5 office tasks you could hand to a dot?

Each of these is recurring, low-judgment until the last step, and ends with a person approving. That is the shape of work dots are designed for.

### 1. Weekly vendor follow-ups

**The problem:** an office manager has a dozen open purchase orders and three vendors who have not confirmed delivery dates. Chasing them is ten minutes each, so it slides to Friday, then to next week.

**The hand-off:** "Every Monday morning, check the open purchase orders sheet. For any order with no confirmed delivery date after 5 business days, draft a short follow-up to the vendor contact and show me the drafts before anything is sent." Set a custom rule of Ask before taking action for sending email (see the safety section below).

### 2. Watching an inbox for invoices

**The problem:** invoices arrive in a shared inbox mixed in with everything else. Someone has to spot them, log them and check for duplicates before the payment run.

**The hand-off:** connect the inbox with read access only, and ask the dot to log each new invoice (vendor, number, amount, due date) into a sheet, flag possible duplicates and post a daily list for review. Never let a dot pay anything. If you want the same job as a fixed automation, the free [invoice inbox to approval sheet lesson](https://learn.ynteractive.com/content/invoice-inbox-to-google-sheets-approval-n8n) builds it in n8n.

### 3. A daily competitor digest posted to Slack

**The problem:** a marketing lead checks four competitors' pricing pages and news by hand, or more often does not check at all, and finds out about a price cut from a customer.

**The hand-off:** "Each weekday at 8 AM, check these four pricing and news pages and post five bullets to #marketing in Slack: what changed, with links. Say 'no change' if nothing did." Treat what websites say as untrusted text, and keep the dot on the pages you named.

### 4. Catching unbilled and overdue invoices

**The problem:** finished work that never gets invoiced, and invoices that age from 30 to 90 days because chasing is awkward. This is exactly OpenAI's launch example.

**The hand-off:** ask the dot to compare completed jobs against invoices sent each Friday, list anything unbilled, and draft polite reminders for invoices more than 14 days late, held for your approval. The free lesson [Collect Overdue Invoices with ChatGPT](https://learn.ynteractive.com/content/collect-overdue-invoices-chatgpt) teaches the chase-list part on any paid plan today.

### 5. Keeping a proposal or client deck current

**The problem:** a sales lead's proposal goes stale every time the customer changes a requirement, and updating it means rereading emails and call notes.

**The hand-off:** OpenAI describes a dot that checks customer requirements against product docs, updates the proposal as requirements change and flags open concerns. Give it the deal folder, ask it to keep a running list of what changed since the last version, and review every change before the proposal goes to the client.

## How do you keep a dot safe at work?

A dot is more useful the more it can reach, and that is also the risk. Set these up before you hand it real work.

**Permissions.** Connect only the apps a task needs. Read access and send access are separate: the help center notes that permission to read information may not include permission to change it or send messages. Leave local computer access off unless you need it.

**Approvals.** Dots start with built-in rules for when to act and when to ask. With Custom rules you choose one of four behaviors for an action: take action without asking, take action if pre-approved, ask before taking action, or hand off to you. Pre-approved means you asked for that action in your prompt, and OpenAI is clear that asking for a draft does not authorize sending it. Actions that could affect your accounts or share information go through an automatic check OpenAI calls auto-review, and some steps, such as changing a password, always stay with you. Start with Ask before taking action for anything that sends, buys or deletes.

**Data.** OpenAI says it does not use content from Business, Enterprise or Edu workspaces to improve its models by default; on personal plans you control this in [ChatGPT data controls](https://learn.ynteractive.com/content/glossary/chatgpt-data-controls) (Settings, then Data controls, then Improve the model for everyone). Two details matter for a business: disconnecting an app does not delete what your dot already learned from it, and the only way to delete a dot's own memories is to reset (delete) the dot. Decide what client data a dot may see before you connect anything, the way the free [ChatGPT data controls lesson](https://learn.ynteractive.com/content/chatgpt-data-controls-protect-client-data) sets it up for a small firm.

**Mistakes and attacks.** OpenAI says dots can still make mistakes, including when following your rules, and tells users to review consequential work. Web pages and emails can contain hidden instructions aimed at AI ([prompt injection](https://learn.ynteractive.com/content/glossary/prompt-injection)); OpenAI says its safeguards watch for this and can pause or stop a dot. Keep a [human in the loop](https://learn.ynteractive.com/content/glossary/human-in-the-loop) for anything a client will see.

**The incidents in the news.** In early October 2026, OpenAI disclosed that one of its AI agents had accessed a New South Wales government web application holding non-public bushfire data back in June, without authorization; reports say no personal information was taken. The reports describe an OpenAI agent, not a customer's dot, but it is a fair reminder that an agent with a browser can go further than intended. Give a dot the narrowest access that does the job, and use the Pause button in its profile when in doubt.

## Can't get dots yet? What can you do today on Plus?

Most of the business value is available now if you are willing to start the work yourself.

1. **Hand off one job with ChatGPT Work, then schedule it.** In the free lesson [ChatGPT Work: Have ChatGPT Do a Multi-Step Task](https://learn.ynteractive.com/content/chatgpt-work-multi-step-task), you have ChatGPT compare three suppliers' prices into a spreadsheet, approve each website it opens, check the file, then ask it to recheck every Monday at 7 AM. That scheduled task is your vendor follow-up and competitor digest without a dot. Plus allows 5 active scheduled tasks.
2. **Give ChatGPT a memory for one area of work with a Project.** The free lesson [ChatGPT Projects: Keep Your Brand Voice, Files and Instructions in One Place](https://learn.ynteractive.com/content/chatgpt-projects-brand-voice) shows how to keep instructions and files so every chat starts with your context, which is the part of a dot that knows your standards.
3. **Learn when an agent is the right tool.** [AI Agent vs Chatbot vs Automation](https://learn.ynteractive.com/content/ai-agent-vs-chatbot-vs-automation) runs one overdue-invoice job three ways, so you can tell which tasks need an agent at all.

## Key takeaways

- A ChatGPT dot is an always-on agent with its own cloud computer that keeps working on a goal between conversations, powered by GPT-6 Astra.
- Access is Pro (not in the EEA, Switzerland or the UK), Business Premium, and an admin-enabled Enterprise beta. The first dot is included in the plan.
- Good first jobs are recurring and end with a person approving: vendor follow-ups, invoice logging, a Slack digest, unbilled invoices, a living proposal.
- Set read-only access and Ask before taking action rules first. Disconnecting an app does not delete what the dot learned.
- On Plus, ChatGPT Work plus a scheduled task and a Project covers much of the same ground today.

## Frequently asked questions

**Are ChatGPT dots available on the Plus plan?**

No. As of October 14, 2026, dots are on Pro, Business Premium and an admin-enabled Enterprise beta. OpenAI says it plans to expand to more users but has not said when.

**What model powers ChatGPT dots?**

OpenAI's announcement and help center say dots are powered by GPT-6 Astra. Some news coverage says GPT-6; we go with OpenAI's wording.

**Do dots cost extra on top of Pro or Business Premium?**

Your first dot is included at no extra cost, and conversations with it do not count toward your usage limits. Deeper work uses an allowance in your plan, and tasks it starts in ChatGPT Work or Codex count toward those limits. OpenAI says more dots and more capacity will be available later.

**Can a dot send emails or pay bills on its own?**

It can only act through the apps and permissions you connect, and Custom rules decide whether it acts, asks first or hands off to you. In proactive research its tools are read-only. For a business, keep sending on Ask before taking action and never give a dot payment access.

**Are dots available in the UK or EU?**

Not on Pro: Pro access excludes the EEA, Switzerland and the UK. OpenAI says Business Premium is available across all supported ChatGPT regions, and Enterprise depends on your admin.

## Sources

- [OpenAI: Introducing dots](https://openai.com/index/introducing-dots/) (September 29, 2026)
- [OpenAI Help Center: Getting started with your dot](https://help.openai.com/en/articles/20001530-getting-started-with-your-dot)
- [ChatGPT pricing](https://chatgpt.com/pricing) and [ChatGPT Business](https://chatgpt.com/business/)
- [TechRadar: OpenAI launches Dots](https://www.techradar.com/pro/openai-launches-dots-its-always-on-ai-agents-that-are-always-watching) (September 30, 2026)
- [Tom's Guide: OpenAI announces ChatGPT Dots](https://www.tomsguide.com/ai/openai-gave-chatgpt-a-virtual-computer-to-run-errands-for-you) (September 29, 2026)
- [Forbes: What Are OpenAI Dots And Do You Need Them?](https://www.forbes.com/sites/ronschmelzer/2026/10/05/what-are-openai-dots-and-do-you-need-them/) (October 5, 2026)
- [OECD AI Incidents Monitor: OpenAI AI agent breaches Australian government websites](https://oecd.ai/en/incidents/2026-10-02-ef44)

## Practice it free

**ChatGPT Work: Have ChatGPT Do a Multi-Step Task**. Switch ChatGPT to Work, hand it a real job (compare three suppliers into a spreadsheet), approve the sites it opens, skip a login, answer its question, check the file and make it a weekly task. You do every step yourself in a practice copy of ChatGPT, it checks your work as you go, and “Do it for me” finishes any step you get stuck on. Free, and nothing touches your real accounts.

[Start the free lesson](https://learn.ynteractive.com/modules/chatgpt-work)

[Or read the full step-by-step guide](https://learn.ynteractive.com/content/chatgpt-work-multi-step-task)

## Free lessons for this

- [ChatGPT Work: Have ChatGPT Do a Multi-Step Task](https://learn.ynteractive.com/content/chatgpt-work-multi-step-task): Switch ChatGPT to Work, hand it a real job (compare three suppliers into a spreadsheet), approve the sites it opens, skip a login, answer its question, check…
- [ChatGPT Projects: Keep Your Brand Voice, Files and Instructions in One Place](https://learn.ynteractive.com/content/chatgpt-projects-brand-voice): Set up a ChatGPT Project with your brand voice instructions, your price list and your style guide, so every new chat already knows your business.
- [AI Agent vs Chatbot vs Automation: One Job Done Three Ways](https://learn.ynteractive.com/content/ai-agent-vs-chatbot-vs-automation): Chase six unpaid invoices three ways: a chatbot writes a template, an automation follows a rule, and an AI agent works out the steps and flags the tricky one.
- [Collect Overdue Invoices Faster using ChatGPT](https://learn.ynteractive.com/content/collect-overdue-invoices-chatgpt): Upload your aging report, have ChatGPT build this week’s chase list (oldest and biggest first, disputes held back), check it against the file, then draft a…
- [ChatGPT: Keep Client Data Safe, Set Up Your Team's Data Controls in 10 Minutes](https://learn.ynteractive.com/content/chatgpt-data-controls-protect-client-data): A client asks if their files go into ChatGPT. Answer with confidence: set the data controls, memory, Temporary Chat and shared links, then hand your team a…

## Related terms

Plain-English definitions, each with an example from a free lesson.

[ChatGPT dot](https://learn.ynteractive.com/content/glossary/chatgpt-dot) · [Always-on agent](https://learn.ynteractive.com/content/glossary/always-on-agent) · [OpenAI](https://learn.ynteractive.com/content/glossary/openai) · [AI agent vs assistant](https://learn.ynteractive.com/content/glossary/ai-agent-vs-assistant) · [AI incident response](https://learn.ynteractive.com/content/glossary/ai-incident-response) · [AI vendor risk assessment](https://learn.ynteractive.com/content/glossary/ai-vendor-risk-assessment) · [Persistent agent](https://learn.ynteractive.com/content/glossary/persistent-agent) · [ChatGPT data controls](https://learn.ynteractive.com/content/glossary/chatgpt-data-controls)

[The full AI glossary](https://learn.ynteractive.com/content/glossary)

## More AI explainers

- [AI Governance for Small Business: A Practical Starter Kit (Free Policy Template)](https://learn.ynteractive.com/content/explainers/ai-governance-for-small-business)

[All AI explainers](https://learn.ynteractive.com/content/explainers) · [All free tutorials](https://learn.ynteractive.com/content)