Always-on agents are finally here. What Dots means for your week.
OpenAI shipped Dots at DevDay. Here is how a small business owner can put one to work on proposals, invoices, and replies.
What shipped
On September 29, 2026, OpenAI announced Dots at DevDay: always-on agents powered by GPT-6 Astra. Per the official announcement, each Dot gets its own cloud computer and browser, learns from feedback over time, and can keep working toward your goals even when your laptop is closed. Through OpenAI’s plugin ecosystem, Dots can connect to over 4,000 apps (vendor claim).
You reach the same Dot on three live surfaces today: ChatGPT (desktop, web, and mobile), Slack, and Microsoft Teams. Context carries across those channels. Texting is listed as coming soon.
When idle, a Dot can do proactive research using already-connected apps with tools restricted to read-only. Per OpenAI, that mode cannot send messages, change app content, or control the browser or computer. Custom Rules let you allow actions, require your approval, or block them. Sensitive tasks such as changing a password stay with you. Activity View shows what it has been doing in the background.
What it means for a business owner
If you already use ChatGPT for proposals, invoices, and customer replies, Dots change the rhythm. You are no longer opening a fresh chat every time you need something done. You hand a Dot a responsibility, it keeps working between conversations, and it comes back when a decision needs you.
Cute avatars are not the story. Permissions are. Before you connect your inbox, bank portal, or customer list, decide what the Dot may draft on its own, what needs your yes, and what it may never send.
How you could use this yourself this week
- 01
Keep one live proposal current.
Share the draft and the customer’s latest notes. Ask your Dot to update seat counts, pricing tables, and open questions as replies arrive, and to bring you a revised draft for approval before anything goes out.
- 02
Catch invoices before they slip.
Point it at the projects or publications you usually forget to bill. Ask it to prepare the invoice and wait for your approval, the pattern OpenAI described from an early tester, rather than sending on its own.
- 03
Draft replies, do not auto-send.
Connect the channel where customer questions pile up. Tell the Dot to draft answers in your voice and stop until you approve. Use Custom Rules so sending always needs you.
- 04
Set the kill switch before you experiment.
Name which apps are connected, which actions need approval, and how you pause the Dot from Activity View. Ten minutes of setup here saves a messy afternoon later.
Access, in plain English (vendor claims)
Dots are rolling out to ChatGPT Pro and Business Premium users in eligible markets. Enterprise, Edu, and Healthcare workspaces can try a beta when a workspace admin enables it. The first Dot is included with Pro or Business Premium at no extra cost (OpenAI claim). Conversations with a Dot do not count toward ChatGPT usage limits; tasks the Dot starts in Codex or ChatGPT Work still count as usual (OpenAI claim). Create the first Dot in the ChatGPT desktop app or desktop browser, then message it from mobile after setup. Rollout is gradual, so an eligible account may not see Dots on day one.
The honest limits
A Dot can still make mistakes. Stopping a task does not undo actions it already completed. Proactive research is read-only by design, but connected plugins and a linked laptop expand what it can touch once you grant write permissions. Treat the first week as shadow mode: watch, approve, correct, then widen the scope.
AI news moves fast, and most of it does not deserve your attention. An always-on agent that can keep a proposal, an invoice queue, and a reply draft moving while you run the business? That one is worth an hour of setup this week.
Want this working in your business?
Tell us what you are trying to fix and we will point you at the smallest useful next step.
Talk to us