OpenAI Dots: Always-On Agents on GPT-6 Astra

Introduction

OpenAI introduced Dots on September 29, 2026, during its DevDay keynote, and describes them as "remarkably capable, always-on agents built to handle everything." A dot is a personal agent that runs on its own cloud computer, connects to apps through plug-ins, and keeps working toward your goals when you are not talking to it. It is powered by GPT-6 Astra. Outlets framed the launch as OpenAI's counterpart to Meta's Muse agent: The Verge headlined it "OpenAI launches Dots, its Muse competitor," and WIRED titled its piece "OpenAI's Dots Are Always-On AI Agents—and Its Answer to Meta's Muse." This review covers only Dots: what OpenAI says they do, where they run, what safeguards come with them, and what remains unanswered.

Feature Overview

A dedicated cloud computer. According to OpenAI's announcement, each dot has "their own cloud computer, their own browser, and the apps you've connected." You can open your dot's computer at any time to inspect its work. A dot can also connect to other devices, and you can give it permission to use your laptop so it works alongside you. The Verge adds that a dot can reach a web browser and more than 4,000 supported apps; OpenAI's own wording is that plug-ins let dots connect to "over 4,000 apps."

Reach across channels. You can message or call a dot in ChatGPT on desktop, web, and mobile, and message it in Slack and Teams. OpenAI says texting is "coming soon." The post says dots carry context across every channel, and that a dot learns your preferences over time from feedback.

Proactive background work. When you are not actively working with it, a dot looks for ways to help, which OpenAI calls "proactive research." That mode uses connected apps through tools restricted to read-only access, so it cannot send messages, change app content, or control your browser or computer. OpenAI's illustrations include a bug appearing in Slack that dots start investigating, and an early tester whose dot noticed an unsent invoice, prepared it, and sent it after his approval.

Specialist dots for organizations. Beyond a personal dot, OpenAI previewed specialist dots that a company sets up with their own identity, credentials, and system access to handle defined responsibilities. OpenAI says it is building on early internal testing across procurement, invoice processing, email marketing, customer support, and commercial contracting, starting with focused enterprise pilots. It is also working with Microsoft to integrate specialist dots with Agent 365 governance and security controls.

Usability Analysis

Setup starts on the desktop. OpenAI says you create your first dot in the ChatGPT desktop app or a desktop browser, connect your apps, and let it introduce itself; after that you can message it from the mobile app. You name your primary dot and customize it. The Verge reports that for now you can only create one dot, although OpenAI says it envisions "teams of dots working together on your behalf" over time.

Usage limits are split in a way worth noting. Conversations with your dot do not count toward ChatGPT usage limits, but tasks a dot starts or manages in Codex or ChatGPT Work count as usual. The plan also includes an allowance for deeper work, with extended limits for the first month after launch. OpenAI says that later you will be able to add more dots and scale each one's speed or monthly workload, but the announcement post does not give prices for those options.

TechCrunch offers a sober framing: much of this functionality was already possible through Codex and similar agent harnesses, and Dots bundles those features into a package focused on independent action, with a cartoonish, bubbly persona similar to Muse. For a developer or an operations lead who already delegates to Codex, the practical change is the persistent identity, the cross-channel context, and the proactive background mode rather than a new class of task.

Safeguards and Control

OpenAI says dots build on ChatGPT protections with additional safety and privacy safeguards. Each dot works on its own cloud computer, and your computer and its contents stay separate unless you choose to connect it. For supported sign-ins, dots can use saved passwords without exposing them to the model. Security safeguards are meant to defend against malicious instructions, and if monitoring detects a safety concern it can pause or stop the dot's work.

Permissions run through existing ChatGPT app controls. Dots start with built-in rules for when to act independently and when to ask for approval, and Custom Rules let you allow, require approval for, or block specific actions. An Activity View shows progress, including background work. Actions that could affect your accounts or share information are checked by auto-review, and certain sensitive tasks, such as changing a password, always stay with the user. OpenAI adds a plain caution: "Dots can still make mistakes, so always review consequential work."

On data, OpenAI says it does not use content from ChatGPT Business, Enterprise, or Edu workspaces to improve its models by default. On personal plans, users can control whether dots' conversations and work are used for training. OpenAI says it does not train directly on proactive research or a dot's notes to itself, though information from them may be used to inform an eligible conversation or task depending on settings.

Pros and Cons

The strengths are a persistent agent with its own inspectable workspace, no charge for the first dot beyond the plan, conversations that do not consume usage limits, and layered permission controls. The limits are equally concrete: availability is restricted to specific paid plans in eligible markets, the personal launch starts with a single dot, texting and multi-dot teams are still future items, and specialist dots are at the pilot stage. Because the agent acts on connected accounts, mistakes have real consequences, which OpenAI itself acknowledges.

Outlook

What OpenAI has committed to in its own post is narrow: a wider rollout "soon," more dots per user later, and deeper enterprise integration through Agent 365. The competitive comparison with Meta's Muse is an outlet framing, and launch-day descriptions alone cannot settle how the two products compare.

Conclusion

Dots repackage agentic capabilities into a named, always-available assistant with its own cloud computer and a clear permission model. They suit Pro and Business Premium users who already delegate work to agents and want continuity across ChatGPT, Slack, and Teams. Buyers should wait for hands-on reports on reliability and for pricing beyond the first dot before treating it as a dependable colleague.

Editor's Verdict

OpenAI Dots: Always-On Agents on GPT-6 Astra earns a solid recommendation within the GPT space.

The strongest case for paying attention: each dot has its own cloud computer and browser that the user can open to inspect its work at any time. That alone raises the bar for what readers should expect in this space. Reinforcing that, the first dot is included in Pro and Business Premium plans at no extra cost, and conversations with it do not count toward ChatGPT usage limits — practical value rather than just headline appeal. The broader signal worth registering is straightforward: the product's core idea is persistence, since a dot keeps working in the background and carries context across ChatGPT, Slack, and Teams rather than resetting each session. On the other side of the ledger, one constraint is real rather than a marketing footnote: availability is limited to specific paid plans in eligible markets, and Enterprise access is a beta that a workspace admin must enable. It should factor into any serious decision. Layered on top of that, the launch offers one personal dot per user according to The Verge, while multiple dots and teams of dots remain future plans — which narrows the set of teams for whom this is an obvious yes.

For ChatGPT power users, OpenAI API customers, and enterprise teams already running on the OpenAI stack, this is a serious evaluation candidate, not just a curiosity to bookmark. For everyone else, the safer posture is to monitor coverage and revisit once the use cases that matter to your team are demonstrated in the wild.