Enterprise
Specialist Dots: Organizational Agents with Their Own Identity
Personal Dots belong to a person. Specialist Dots belong to an organization — with their own identity, credentials, and access to the systems of record.
Last verified: September 30, 2026 · Reviewed by AstraDot editorial desk
The Dots most people meet first are personal: you start with a primary Dot, name it, and give it rules. But OpenAI is also building a second, organizational kind of Dot for companies. These are called specialist Dots, and they change the governance conversation because they act as the organization rather than as an individual. If you are evaluating Dots for a larger team, this is the concept to understand before you commit.
This page explains what a specialist Dot is, how it differs from a personal Dot, what OpenAI is testing, and how to think about oversight. It builds on the general explainer for what OpenAI Dots are and the security and privacy guide.
What a specialist Dot is
A specialist Dot is an agent that takes on a dedicated responsibility inside an organization. Rather than being a general assistant that answers whatever you ask, it is adapted to a role and given the access that role needs. OpenAI describes specialist Dots as having their own identity, credentials, and access to the systems of record — the authoritative systems where a business keeps its data, such as its finance or customer platforms. It improves with team feedback, the same way your personal Dot learns from yours.
How it differs from your personal Dot
The distinction is less about raw capability and more about who the agent represents. A personal Dot is an extension of one person's work. A specialist Dot is a piece of organizational infrastructure, with an identity that can be governed like any other service account.
| Aspect | Your personal Dot | A specialist Dot |
|---|---|---|
| Who it belongs to | You | The organization |
| Identity | Your primary Dot, named by you | Its own identity and credentials |
| Systems access | Apps you connect for your own work | Access to the systems of record it needs |
| How it improves | Learns from your feedback over time | Role-adapted, improves with team feedback |
| Oversight | Your Custom Rules and activity view | Organizational reviews, approvals, and governance |
Because a specialist Dot touches shared systems, its access is reviewed and approved by the organization rather than configured by a single user. That is a deliberate trade-off: more capability, but also more process around who may grant and revoke it.
The functions OpenAI is testing
OpenAI is exploring specialist Dots in specific business functions rather than launching a generic "company agent." The areas it names are:
Procurement
Sourcing and purchasing workflows inside an organization.
Invoice processing
Handling invoices against the systems of record.
Email marketing
Campaign work tied to the organization's marketing stack.
Customer support
Support workflows with access to the relevant systems.
Commercial contracting
Contract-related work in the commercial pipeline.
These functions share a pattern: each has well-defined inputs, authoritative systems of record, and outcomes that can be checked. That makes them sensible places to test an agent that must act with an organization's own credentials.
How the pilots work
OpenAI says it is starting with focused enterprise pilots and working directly with organizations. Practically, that means the early specialist Dots are not a self-serve product you can switch on from a pricing page. If your organization is a fit, the path is a partnership conversation rather than a checkout flow. Timings, eligibility, and pricing for specialist Dots are not published, so do not anchor a business case to a date that has not been confirmed. Watch OpenAI's announcement and help center for updates.
Microsoft Agent 365 integration
For many enterprises the deciding question is not capability but control. OpenAI is working with Microsoft to integrate specialist Dots with enterprise governance and security controls in Agent 365. That is significant: it means organizations expect to manage specialist agents where they already manage everything else — identity, policy, and oversight. If you are comparing the two approaches directly, our Dots vs Microsoft Agent 365 page breaks down the overlap and the differences.
How to think about governance
A specialist Dot is closer to a new hire with standing credentials than to a chatbot, so the governance questions are correspondingly familiar:
- Least privilege. Give the Dot the narrowest access the role needs, and revisit it as the role changes.
- Clear ownership. Every specialist Dot should have a named human owner accountable for its behaviour.
- Review before consequence. Keep the actions that move money, data, or public messages behind organizational approvals.
- Auditability. Use the activity view and your existing logging so decisions can be reconstructed later.
- Human judgement stays in the loop. Dots can make mistakes, so consequential output should be reviewed by a person.
None of this is unique to Dots — it is the same discipline you would apply to any service with privileged access. The novelty is that the "service" is now an agent that can reason and act, which is exactly why the governance deserves to be set up before the pilot, not after.
Source: OpenAI, Introducing dots, checked 2026-09-30. Items OpenAI has not published are marked "not published" rather than estimated.