ChatGPT Dots Explained: OpenAI’s New Always-On AI Agents, Features and Benefits
Published: September 30, 2026 | Updated: September 30, 2026 | By CodeMicros
Quick answer: ChatGPT Dots is OpenAI’s new always-on AI-agent experience. A “dot” can work toward ongoing goals using its own cloud computer, browser and connected apps, then bring you work, questions or approval requests instead of waiting for a new prompt every time.
OpenAI announced Dots on September 29, 2026. The name can sound like “ChatGPT dot new,” but the product is called ChatGPT Dots. It represents a change from one-off AI chats to an agent that can retain context around a goal, work in the background and return when it has something useful for you to review.
What are ChatGPT Dots?
A dot is an AI agent designed for work that continues after the first instruction. Instead of repeatedly opening separate chats for research, follow-ups, drafts and updates, you give a dot a goal and feedback. It can then keep the surrounding context in mind, look for the next useful step and ask you to approve actions that matter.
According to OpenAI’s announcement, each dot has its own cloud computer and browser. It can use connected apps and can be reached through ChatGPT on desktop, web and mobile, with Slack and Teams support also described by OpenAI. This is important: Dots is not simply a new ChatGPT model name. It is a product experience built around persistent, agent-style work.
How ChatGPT Dots work
The basic workflow is simple:
- Create a primary dot and explain your role, goals, preferences and boundaries.
- Connect only the apps and information sources you want it to use.
- Give it an outcome, such as monitoring customer feedback, preparing launch material or researching a topic.
- Review its progress in Activity View, answer questions and approve sensitive actions when required.
- Improve the results over time by giving clear feedback about what “good” looks like for you.
OpenAI describes proactive background research as read-only when you are not actively working with a dot. In other words, it can inspect approved sources for useful updates, but it should not independently send messages or change content through that read-only mode. Actions that could affect an account or share information are subject to review and approval rules.
Key advantages of ChatGPT Dots
1. Less repetitive prompting
Normal AI chat is excellent for a single request, but project work often needs context repeated many times. A dot is designed to learn your standards and continue a workflow. That can reduce the time spent re-explaining a product, audience, writing style or definition of done.
2. Background follow-through
The most useful difference is follow-through. A dot can watch an approved source, prepare a research update, organize open questions or return with a draft while you focus on another task. For busy founders and developers, that can turn AI from a question-answering tool into an assistant for longer-running work.
3. Work across tools and channels
Dots are intended to work with connected apps and communicate through more than one place. A project can begin in ChatGPT, continue with team context in Slack, and still preserve the goal and review history. This can reduce fragmented notes and missed follow-ups.
4. A practical developer workflow
OpenAI’s example for developers is especially relevant: a dot can examine recurring customer feedback, scope smaller fixes, build and test changes, then present pull requests and evidence for review. The benefit is not “ship code without thinking.” It is reducing the repetitive discovery, triage and preparation work before a developer makes the final engineering decision.
5. Better continuity for creators and small teams
A content creator could use a dot to turn an approved interview transcript into draft show notes, short-clip ideas and social posts. A product team could use it to update launch drafts after a requirement changes. These are useful because the agent is meant to carry feedback from one revision into the next rather than treating every request as a fresh task.
How CodeMicros readers can use Dots responsibly
For a Flutter developer or indie app creator, the most practical uses are structured and reviewable:
- Collect recurring app-review and support feedback, then prepare a prioritised issue brief.
- Watch release notes from approved platforms and prepare a “what changed for our app” summary.
- Turn a feature brief into a QA checklist, acceptance criteria and draft release notes.
- Prepare content outlines, keyword research notes and social drafts for your final review.
- Keep a launch checklist current when screenshots, store metadata or requirements change.
Give the dot a narrow first responsibility. For example: “Review new AroundMe feedback once daily, group it by bug, feature request or confusion, and prepare a report. Do not reply to customers or modify anything without approval.” Specific boundaries make an agent more useful and easier to supervise.
Privacy, approvals and user control
Persistent agents need more care than a normal chat because they may handle connected information over time. OpenAI says a dot’s cloud computer is separate from your own computer unless you choose to connect yours. You choose app access and can manage permissions through ChatGPT controls.
OpenAI also says Dots uses Auto-review to evaluate actions against your instructions, custom rules and safety requirements. Sensitive tasks such as changing a password remain with the user. These safeguards are valuable, but they are not a reason to skip review. Always inspect consequential work: code changes, customer communications, payments, account permissions, confidential documents and published content.
For Business, Enterprise and Edu workspaces, OpenAI says content is not used to improve models by default. On personal plans, users can control whether conversations and work are used to improve models. Check your current settings and your organisation’s data policy before connecting business systems.
Availability, price and important limitations
Dots began rolling out on September 29, 2026 to ChatGPT Pro and Business Premium users in eligible markets. Enterprise users, including Edu and Healthcare workspaces, can try the beta when an administrator enables it. OpenAI says the first dot is included for Pro and Business Premium users at no extra cost, while plans include an allowance for deeper work. Availability, limits and eligible markets can change, so verify the current status in ChatGPT before promising access to a team.
There are sensible limits to remember:
- Dots can make mistakes, so they do not replace human judgment.
- Connected-app permissions should be minimal and reviewed regularly.
- Background read-only research is different from permission to take external actions.
- Tasks started in Codex or ChatGPT Work can count toward those products’ usual usage limits.
- Do not give an agent unrestricted access to sensitive accounts simply for convenience.
How to get started with ChatGPT Dots
- Confirm that Dots is available on your eligible ChatGPT plan and market.
- Create your first dot from the ChatGPT desktop app or desktop browser.
- Name it around a single job, such as “App Release Assistant” or “Content Research Assistant.”
- Write boundaries first: what it may read, what requires approval, and what it must never do.
- Connect the smallest number of apps needed for the task.
- Start with a low-risk workflow and review its output before expanding access.
Final thoughts
ChatGPT Dots is significant because it aims to make AI more continuous and accountable to a goal, not just responsive to the latest message. The biggest advantage is not automatic action for its own sake; it is getting prepared, contextual work back while you keep control of decisions that matter. Used with clear permissions and regular review, Dots could be particularly useful for developers, creators and small teams managing many moving parts.
Frequently asked questions
Is ChatGPT Dots the same as a new GPT model?
No. Dots is an agent experience powered by GPT-6 Astra, rather than a standalone model-picker label.
Is ChatGPT Dots available to every user?
No. At launch, OpenAI says the rollout is for Pro and Business Premium users in eligible markets, with Enterprise beta access controlled by workspace administrators.
Can a dot send messages or change data on its own?
OpenAI describes proactive background research as read-only. Other actions are governed by permissions, review and approval rules; users should always set clear boundaries.
Should developers let a dot merge code automatically?
Not as a starting point. Let it prepare analysis, tests and pull requests first, then keep human review for code quality, security and production decisions.
