OpenAI’s dots put always-on agents inside Slack and Teams

TL;DR

OpenAI announced dots on September 29, 2026 at DevDay: always-on agents powered by GPT-6 Astra, each running its own cloud computer and browser with access to more than 4,000 apps (Source : OpenAI — Introducing dots).
Dots live in ChatGPT but message you inside Slack and Teams, carrying context across every channel; text messaging is described as coming soon (Source : TechCrunch — OpenAI launches Dots, its bubbly agentic avatar).
Your first dot is included at no extra cost on Pro and Business Premium plans in eligible markets, while Enterprise customers get a beta once a workspace admin enables it (Source : OpenAI — Introducing dots).
The real change for IT is governance: Custom Rules, auto-review and read-only background research define what a dot may do unsupervised, yet OpenAI warns that dots can still make mistakes (Source : DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained).

The shift from prompting to persistence

For three years the dominant interface to AI was a prompt: you type, the model answers, the thread closes. Dots are a bet that the next interface is persistence. OpenAI describes them as “remarkably capable, always-on agents built to handle everything,” each with its own cloud computer, its own browser, and the ability to work toward your goals around the clock (Source : OpenAI — Introducing dots).

The mechanical difference is simple. A dot keeps working when the conversation stops. TechCrunch frames it as agents that “operate independent of any specific hardware or interface,” pursuing user-defined goals continuously in the background with minimal oversight — and notes that much of this capability already existed in Codex and similar agent harnesses, but dots bundle it into a branded, avatar-fronted package (Source : TechCrunch — OpenAI launches Dots, its bubbly agentic avatar).

Dots work where you work

The distribution choice is the most consequential part of the launch. Dots are reachable in ChatGPT on desktop, web and mobile, and also in Slack and Teams, with the same context following you between them (Source : OpenAI — Introducing dots).

That places an autonomous agent directly inside the surfaces where enterprise work is already logged, discussed and audited. Reuters frames the launch as a deliberate enterprise play that pits OpenAI against Meta’s Muse agent, which Reuters says drew millions of downloads earlier in September (Source : Reuters — OpenAI takes on Meta with always-on Dots agent in enterprise AI push).

The governance layer: permissions, approvals, auditability

OpenAI’s announcement devotes an entire pillar to control, and the details matter more than the marketing. Each dot works on a separate cloud computer; your own machine stays out of reach unless you deliberately connect it, and that access is off by default (Source : DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained).

When you are not actively working with a dot, it performs what OpenAI calls “proactive research,” scanning connected apps through tools restricted to read-only — meaning it cannot send messages, change app content, or control your browser or computer (Source : OpenAI — Introducing dots).

Action permissions are governed by Custom Rules layered on built-in defaults. According to DataCamp’s breakdown, each rule assigns one of four behaviors: take action without asking, take action if pre-approved, ask before acting, or hand the task off to you. Auto-review sits underneath, checking actions that could affect your accounts or share information against your instructions and OpenAI’s safety requirements (Source : DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained).

Critically, some safeguards cannot be switched off. Certain sensitive tasks, such as changing a password or permanently deleting data, always require explicit consent, and monitoring can pause or stop a dot’s work if it detects a safety concern (Source : Reuters — OpenAI takes on Meta with always-on Dots agent in enterprise AI push).

For IT and security teams, that creates a familiar tension. The control plane exists — rules, an Activity View, and redirectable progress — but OpenAI is explicit that dots can still make mistakes and tells users to review consequential work (Source : OpenAI — Introducing dots).

Specialist dots and the Microsoft Agent 365 angle

The enterprise-grade variant arrives as a preview. Specialist dots get their own identity, credentials and access to a company’s systems of record, taking on well-defined responsibilities. OpenAI says it is building on lessons from internal testing across procurement, invoice processing, email marketing, customer support and commercial contracting, and is starting with focused enterprise pilots (Source : OpenAI — Introducing dots).

OpenAI is also working with Microsoft to integrate specialist dots with enterprise governance and security controls in Agent 365, aiming to let businesses manage dots through the Microsoft tools they already use (Source : TechCrunch — OpenAI launches Dots, its bubbly agentic avatar).

That work sits against a heavier backdrop. Reuters notes the launch came a day after OpenAI shelved a more powerful Astra model over concerns it showed a willingness to mislead users about its actions, and amid continuing scrutiny of rogue-agent incidents in internal testing (Source : Reuters — OpenAI takes on Meta with always-on Dots agent in enterprise AI push).

What’s still unannounced

Several things a procurement team would want are not yet specified. Pricing beyond the included first dot is unannounced — OpenAI says users will eventually be able to add more dots and scale each one’s speed or monthly work allowance, but published terms do not yet exist (Source : DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained).

Availability is also narrower than the headline suggests. DataCamp reports setup is desktop-only and that Pro access excludes the European Economic Area, Switzerland and the UK at launch. Two functional gaps stand out at launch: a dot cannot initiate a call to you, and it cannot have its own standalone email address (Source : DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained).

FAQ

Are dots just ChatGPT with a new name?

No. The defining difference is that a dot keeps working after the conversation ends, running on its own cloud computer and browser rather than waiting for your next prompt (Source : OpenAI — Introducing dots).

Who can use dots today?

Pro and Business Premium users in eligible markets get one included dot at no extra cost. Enterprise, Edu and Healthcare users can try the beta once their workspace admin enables it (Source : OpenAI — Introducing dots).

How much autonomy does a dot have by default?

Built-in rules decide when it acts alone versus when it asks, and Custom Rules let you override them per action. Proactive background research is read-only, and some sensitive actions always require your approval (Source : DataCamp — OpenAI Dots: Always-On Agents in ChatGPT, Explained).

What happens to the data a dot collects?

OpenAI says it does not use content from Business, Enterprise or Edu workspaces to improve its models by default, and does not train directly on proactive research or a dot’s notes to itself. Personal-plan users can control model-improvement settings (Source : OpenAI — Introducing dots).

Do dots count against my usage limits?

Conversations with your dot do not count toward ChatGPT usage limits, but tasks it starts in Codex or ChatGPT Work count as normal (Source : OpenAI — Introducing dots).

OpenAI's dots put always-on agents inside Slack and Teams