OpenAI launches Dots, always-on GPT-6 Astra agents on Pro and Business Premium that carry work across Slack, Teams and ChatGPT

An AI assistant that only answers in a chat box can only do the work a person asks about in that chat, and OpenAI's answer is a new agent that opens its own apps and does work while nobody is watching it.

OpenAI announced Dots at its DevDay keynote on Tuesday 29 September. Dots are always-on agents powered by GPT-6 Astra, each with its own cloud computer, its own browser, and access to more than 4,000 apps through OpenAI's plugin ecosystem. The company says a Dot "can do nearly anything using their own cloud computer, their own browser, and the apps you've connected", and that its ecosystem of plugins lets a Dot "readily connect to over 4,000 apps".

A cloud computer per Dot, running while nobody looks

Each Dot runs on a cloud computer that is separate from the user's own machine. You can open that cloud computer to inspect what the Dot is doing, and you can also let the Dot connect to your laptop when you want it working next to you. OpenAI says a Dot can take on a project and keep several running at once. The examples in the launch post include a Dot investigating a bug in Slack, turning a design into a working app, and preparing an invoice a user had forgotten to send.

You reach a Dot in ChatGPT on desktop, web and mobile, by message or voice call. It can also work inside Slack and Microsoft Teams, and OpenAI says text-message support is coming soon. Context follows the Dot from one channel to the next.

The safety model is a set of rules

Each Dot ships with "built-in rules for when to act independently and when to ask for approval", and users can add Custom Rules that block a specific action, require approval, or allow one outright. Dots use an auto-review system that checks each action against those rules and OpenAI's safety requirements before it runs, and OpenAI says certain sensitive tasks, such as changing a password, "always stay with you". A monitoring system can pause or stop a Dot if it detects a safety concern. During background "proactive research", connected apps are restricted to read-only, so a Dot cannot send messages or change content while the user is away.

Included in Pro and Business Premium, with a first-month capacity bump

Dots are rolling out today to ChatGPT Pro and Business Premium users in eligible markets, and Enterprise, Edu and Healthcare workspaces can turn on the beta from their admin console. The first Dot is included in the Pro or Business Premium plan at no extra cost, and OpenAI is offering extended limits for the first month after launch. Conversations with a Dot do not count against ChatGPT usage limits, but tasks the Dot hands off to Codex or ChatGPT Work do. OpenAI says future updates will let users add more Dots and pay to raise the speed or monthly work capacity of each one.

OpenAI is also previewing "specialist Dots" for enterprises, with their own identities, credentials and access to systems of record. The company says its engineering teams will work directly with pilot organisations, and that it is building an integration with Microsoft's Agent 365 governance and security controls so businesses can manage specialist Dots through the Microsoft tooling they already use.

The Verge and TechCrunch both frame Dots as OpenAI's answer to Meta's Muse agent, which launched weeks earlier and uses a similar chat interface. Dots have their own cartoon avatars and each one can be renamed by its user.

Source

• Primary: OpenAI, "Introducing dots"
• Reporting: Emma Roth, The Verge, "OpenAI launches Dots, its Muse competitor"; Rebecca Bellan, TechCrunch, "OpenAI launches Dots, its bubbly agentic avatar"