OpenAI is beginning to roll out Dots, persistent agents in ChatGPT that can work across connected apps and continue projects in the background. The company’s launch announcement describes them as powered by GPT-6 Astra, with their own cloud computers and access to more than 4,000 apps through plugins.
The rollout starts with Pro and Business Premium users in eligible markets. Enterprise users, including Edu and Healthcare workspaces, can try a beta when their workspace admin enables it.
A computer of its own
OpenAI says a Dot can use its own browser and connected apps, take on several projects at once, and learn from feedback. Users can inspect its cloud computer or optionally give it access to another device, including their laptop.
Dots can be reached through ChatGPT on desktop, web and mobile, as well as Slack and Teams. Texting is planned, rather than available as part of the announcement. Setup starts in the ChatGPT desktop app or a desktop browser; afterward, users can message their Dot from the mobile app.
Those are OpenAI’s descriptions of the product’s capabilities. The company also warns that Dots can make mistakes and tells users to review consequential work.
What can happen without approval?
Users choose which apps to connect. OpenAI says Custom Rules can allow specific actions, require approval or block them, while built-in safety requirements continue to apply. Activity View lets users follow background work and redirect the agent.
The announcement draws a distinction between acting on a task and looking for ways to help. When a user is not actively working with a Dot, its “proactive research” uses read-only tools in already connected apps. OpenAI says those tools cannot send messages, change app content or control a browser or computer.
