OpenAI announced Dots at its DevDay 2026 event, an always-on AI agent feature built into ChatGPT. Dots are described by the company as agents with their own cloud computer that work 24/7 on tasks you assign, learn from your feedback over time, and connect to more than 4,000 apps through OpenAI's plugin ecosystem. The launch drew simultaneous coverage from 9to5Google, 9to5Mac, Notebookcheck and other outlets.
The announcement lands as competition heats up in the consumer AI agent space, following Meta's Muse avatar agent and SpaceXAI's Grok Bot. OpenAI positions Dots as a more capable, background-running alternative, powered by its new GPT-6 Astra model. Dots are available today to Pro and Business Premium subscribers, with Enterprise users able to try them once workspace admins approve.
How OpenAI Dots work
According to OpenAI, each Dot has its own cloud computer and can be given tasks that run in the background. You can watch the Dot's cloud computer at any time while it works, and you can grant it permission to use your own computer. Interaction happens through ChatGPT, Slack or Teams, and a voice call option is available when you need to talk something out.
OpenAI says Dots follow the same permissions and controls as ChatGPT, and support custom rules about when they can act on their own or require your approval. Examples shared by the company include a Dot noticing an invoice had not been sent and preparing it for approval, and Dots automatically investigating bugs when they are mentioned in Slack. When not working on an assigned task, Dots perform what OpenAI calls proactive research through read-only connections to your tools.
- Always-on agents with their own cloud computer
- Powered by GPT-6 Astra
