OpenAI has launched dots, a new class of always-on AI agents designed to keep working toward a user’s goals across apps, devices and communication channels without requiring constant supervision.

Each dot gets its own cloud computer

Dots are powered by GPT-6 Astra and run on dedicated cloud computers. OpenAI says a dot can use its own browser, work across connected applications and continue projects in the background while the user moves on to something else.

Users can inspect the dot’s computer at any time and, if they choose, give it permission to connect to a personal laptop. OpenAI says dots can also carry multiple projects at once instead of forcing users to manage separate conversation threads.

More than 4,000 apps can be connected

Through OpenAI’s plugin ecosystem, dots can connect to more than 4,000 applications. They can also be reached through ChatGPT, Slack and Microsoft Teams, with texting support planned later.

OpenAI says a dot carries context across those channels, so a project started in ChatGPT can continue in Slack or Teams without resetting the conversation.

OpenAI wants dots to become proactive

The core idea is that users should not have to ask for every step. OpenAI describes dots as agents that learn preferences, standards and working style over time, then use that context to prepare work before the user explicitly requests it.

Examples shared by OpenAI include watching customer feedback, preparing tested bug fixes, revising launch materials when product scope changes, rerunning scientific analyses as new data arrives and preparing invoices for approval.

Safety is built around permissions and review

OpenAI says proactive background work is read-only by default. Dots can inspect connected apps, but cannot send messages, change app content or control a browser or computer unless the required permissions are granted.

Users can create Custom Rules that allow, block or require approval for specific actions. Sensitive actions such as changing a password remain reserved for the user. OpenAI also says monitoring systems can pause or stop a dot if potentially harmful behavior is detected.

Specialist dots are coming for companies

OpenAI is also previewing specialist dots that organizations can provision with their own identities, credentials and system access. These agents are intended to take on defined responsibilities such as procurement, invoice processing, support or commercial contracting.

Microsoft and OpenAI are working to integrate specialist dots with Microsoft Agent 365 governance and security controls.

Availability

Dots are rolling out to ChatGPT Pro and Business Premium users in eligible markets, while Enterprise, Edu and Healthcare workspaces can enable a beta through administrators. The first dot is included in Pro and Business Premium plans at no additional cost.

Practical significance

Dots mark OpenAI’s clearest move yet from chat-based assistants toward persistent software that can own parts of a workflow over time. The important shift is not simply that the model can use tools, but that the agent is designed to remember context, stay active and continue working even when the user is not directly interacting with it.

OpenAI announced dots in its official product post .

Source: OpenAI