Specialist dots are the part of the announcement aimed at organisations rather than individuals, and they are a genuinely different thing from the dot you create for yourself. Your own dot works on your behalf. A specialist dot takes on a defined responsibility inside a company, with an identity of its own. Here is what has actually been confirmed, and what has not.
Updated October 2026. Dots launched on 29 September 2026 and are rolling out gradually, so check OpenAI own pages for the current position before relying on any detail here. Specialist dots are described as a preview with focused enterprise pilots, so most details are not yet public.
1. They have their own identity and credentials
This is the defining difference. OpenAI says a company sets up each specialist dot with its own identity, credentials and access to the systems it needs. An agent with its own identity can be granted and revoked like a member of staff, which is what makes it auditable; an agent borrowing a person account cannot.
2. They take on a defined responsibility
The framing is deliberately narrow. Where a personal dot is open-ended, a specialist dot is set up around well-defined responsibilities, with engineering teams working directly with the organisation to define what it is responsible for, which tools it may use, and how people review and approve its work.
3. The early areas are back-office processes
OpenAI says it is building on lessons from early testing inside its own company across procurement, invoice processing, email marketing, customer support and commercial contracting. These are rule-bound processes with clear inputs and an audit trail, which is where this class of agent is easiest to supervise.
4. They are meant to be provisioned like hardware
The preview mentions IT-provisioned hardware and support for deep integrations with a company systems of record. The implication is that a specialist dot is managed as an asset with a lifecycle rather than a chat session, which is the same direction the rest of the industry is moving in.
5. Microsoft Agent 365 is the governance layer
OpenAI says it is working with Microsoft to integrate specialist dots with the enterprise governance and security controls in Agent 365, with the goal of letting businesses manage dots through the Microsoft tools they already use.
Microsoft describes Agent 365 as a control plane that lets IT and security teams observe, govern and secure every agent in an organisation, with a registry in the Microsoft 365 admin center and enforcement through Entra, Intune, Purview and Defender. It has been generally available for the commercial segment since 1 May 2026, and Microsoft documentation says it works best with Microsoft E5 as a prerequisite and that at least one user must hold a qualifying licence to enable it.
On price, Microsoft lists Agent 365 at 15.00 US dollars per user per month paid yearly on an annual commitment, and includes it in Microsoft 365 E7 at 99.00 US dollars per user per month on the same terms. It is licensed per user, so the cost scales with the people governing agents rather than with the number of agents.
6. What has not been published
Being clear about the gaps matters more than usual here, because this is a preview:
- No price for specialist dots, and no indication of how they are licensed.
- No general availability date. OpenAI says it is starting with focused enterprise pilots.
- No published limit on how many a company may run, or how they are grouped.
- No detail on how a specialist dot identity maps onto existing directory systems beyond the Agent 365 work.
For the governance questions this raises in general, see our explainers on AI governance and the EU AI Act. For how a single agent differs from a team of them, see multi-agent systems.
How specialist dots differ from your own dot
The two are easy to conflate because they share a name and a model. The differences that matter are about who the agent belongs to and who answers for what it does:
- Identity. Your dot acts as you, with your access. A specialist dot has credentials of its own.
- Scope. Yours is open-ended and learns your preferences. A specialist dot is set up around a defined responsibility.
- Setup. You create yours in the desktop app in a few minutes. A specialist dot is configured with OpenAI engineering teams during a pilot.
- Oversight. Yours is reviewed by you. A specialist dot is reviewed by a team, and the intent is that it is governed through enterprise tooling.
- Availability. Yours is rolling out now on eligible plans. Specialist dots are a preview.
Why identity is the part that matters
It is tempting to read the identity point as an administrative detail. It is the whole design. An agent that borrows a person account is indistinguishable from that person in every log it touches, which makes it impossible to answer the two questions any audit will ask: what did the agent do, and who approved it doing that.
Giving the agent its own credentials makes both answerable, and it makes the agent revocable without disturbing a member of staff. It also turns agent sprawl into a tractable problem, which is the premise behind Microsoft describing Agent 365 as a registry and a control plane: you cannot govern what you cannot enumerate.
The same logic explains why this is arriving as a pilot rather than a toggle. Provisioning an identity, scoping its access to systems of record, and defining who reviews its work are decisions an organisation has to make deliberately, one process at a time.
What to settle before asking for a pilot
If the idea is appealing, the useful preparation is not technical. It is deciding which single process you would hand over, who owns the output when the agent produces it, and what evidence you would need to believe it is working. OpenAI says its engineering teams work with organisations to define each responsibility, the tools the dot may use, and how people review and approve its work, so arriving with those three answers is what makes the conversation productive.
Common questions
How are specialist dots different from a personal dot? A personal dot works on your behalf using your access. A specialist dot has its own identity, credentials and access, and takes on a defined responsibility inside the organisation.
Can I buy specialist dots today? Not as a self-serve product. OpenAI describes focused enterprise pilots with its engineering teams involved in the setup.
What does Microsoft Agent 365 do? Microsoft describes it as a control plane for discovering, governing, securing and managing agents, with a registry in the Microsoft 365 admin center and enforcement through Entra, Purview and Defender.
What does Agent 365 cost? Microsoft lists it at 15.00 US dollars per user per month paid yearly on an annual commitment, and includes it in Microsoft 365 E7 at 99.00 US dollars per user per month on the same terms.
Which jobs are specialist dots aimed at first? OpenAI names procurement, invoice processing, email marketing, customer support and commercial contracting as the areas it tested internally.
Sources and further reading
Where the figures and rules above come from, so you can check them:
- The announcement, including proactive research and the control model: OpenAI, Introducing dots
- The governance layer OpenAI is integrating specialist dots with: Microsoft, Agent 365
- The admin roles, pillars and licensing for Agent 365: Microsoft Learn, Agent 365 overview
- Everything else announced alongside dots: OpenAI, DevDay 2026 Recap
Photo credits: BioAffix Cardreader Wall by Sezeruludag, CC BY 4.0, via Wikimedia Commons. Cummins Corporate Office Building – April 2008 – Holly Higgins 06 by Holly Higgins, CC BY 2.0, via Wikimedia Commons. Horasis Global China Business Meeting 2013 – boardroom panel (10865474346) by Richter Frank-Jurgen, CC BY-SA 2.0, via Wikimedia Commons.
Join the discussion