OpenAI Dots are a new type of always-on AI agent designed to work toward a user's goals instead of waiting for a prompt every time. OpenAI introduced Dots at DevDay on September 29, 2026, describing them as agents that can run on their own cloud computer, connect to more than 4,000 apps, learn from feedback, and continue working in the background. They are initially rolling out to Pro and Business Premium users in eligible markets, while Enterprise users can access a beta when their workspace administrator enables it. OpenAI's official Dots announcement provides the current product details.
The important distinction is how Dots change the interaction model. A normal chatbot generally waits for a request and returns an answer. A Dot can take a broader objective, work across connected applications, continue a project over time, and ask for approval when an action requires it. Early testers have already reported uses such as email triage, schedule conflict detection, document preparation, and routine office work, although they also reported bugs and permission problems in early builds.
What Are OpenAI Dots?
OpenAI Dots are persistent AI agents that can work on tasks in the background using their own cloud computer, browser, and connected applications. OpenAI says Dots are powered by GPT-6 Astra and can work toward goals continuously, rather than requiring users to direct each individual step. They can be reached through ChatGPT and can also communicate through Slack and Microsoft Teams.
This makes Dots different from a simple chatbot or a fixed automation workflow. The user can give a Dot a project and continue adding context while the agent works through multiple tasks. OpenAI also says Dots can learn a user's preferences and standards from feedback over time.
The product sits in a broader shift toward agentic AI. ToolJunction's existing guide on building AI agents for business operations covers the underlying distinction between systems that simply respond and agents that can use tools, make decisions within defined boundaries, and carry work forward. Dots apply that model as a packaged OpenAI product rather than a framework developers assemble themselves.
How OpenAI Dots Work
Cloud-based execution: Each Dot gets its own cloud computer and browser environment. OpenAI says users can inspect the Dot's computer while it is working and can optionally give it permission to connect to their own device.
Connected applications: Dots can work with applications the user connects. OpenAI says its plugin ecosystem gives Dots access to more than 4,000 apps. The agent can also maintain context across ChatGPT, Slack, and Teams.
Persistent work: Instead of completing one request and stopping, a Dot can keep working on a project, manage several tasks, and return with progress, questions, or decisions that need human input.
Human approval: Dots are not designed to operate without limits. Users choose which applications a Dot can access, define custom rules, and can require approval for specific actions. OpenAI says its auto-review system evaluates actions that could affect accounts or share information.
Key OpenAI Dots Features
| Feature | What It Means | Practical Value |
|---|---|---|
| Always-on operation | Dots can continue working toward tasks in the background | Useful for recurring research and ongoing projects |
| Cloud computer | Each Dot has its own computing and browser environment | Allows work across web-based tools without relying entirely on your local machine |
| 4,000+ app connections | Dots can use connected applications through OpenAI's ecosystem | Useful for cross-application workflows |
| Persistent context | Dots retain knowledge of goals, preferences, and project context | Reduces the need to repeatedly explain a workflow |
| Permissions and approvals | Users can control access and require approval for sensitive actions | Provides a control layer for agentic workflows |
OpenAI Dots Use Cases
OpenAI's examples span software development, product launches, research, sales, and content production. The common pattern is not simply content generation. It is an ongoing workflow where the agent receives new information, updates its work, and returns something for review.
Software development
OpenAI describes a Dot that monitors customer feedback, identifies recurring requests, scopes smaller fixes, builds and tests them, and prepares pull requests for a developer to review. That positions Dots closer to an ongoing engineering assistant than a coding chatbot that only writes code when prompted.
Research and analysis
A Dot can monitor incoming information, rerun analyses when new evidence appears, investigate unexpected results, and update supporting material. This can be useful for teams that repeatedly process new data against the same research question.
Sales operations
OpenAI also describes a sales scenario in which a Dot checks customer requirements and account history, identifies missing technical tests, prepares a proof of concept, and updates a proposal as requirements change. That illustrates the value of persistent context in complex sales cycles.
Content production
For content teams, OpenAI says a Dot can process interview transcripts, identify moments suitable for clips, prepare show notes, and draft social posts for approval. It can carry user edits across those materials and adapt to preferences over time.
Routine administration
Early tester reports provide a different view of the product's practical use. One tester used a Dot connected to email and Slack to filter incoming requests and flag a meeting conflict after a flight was rebooked. Another reported using a Dot for administrative tasks such as responding to a bookkeeper, declining an invitation, and preparing a meeting agenda. These examples are based on early-user reports rather than an independent benchmark.
OpenAI Dots for Businesses
Business deployment is one of the more important parts of the Dots announcement. OpenAI is previewing specialist Dots designed around specific organizational responsibilities. These agents can have their own identity, credentials, and access to the systems required for a defined role.
OpenAI says internal testing has included areas such as procurement, invoice processing, email marketing, customer support, and commercial contracting. The company is starting focused enterprise pilots and says Microsoft Agent 365 integration is being developed so organizations can manage specialist Dots through Microsoft's existing governance and security controls.
That could make Dots relevant to organizations that want persistent digital workers without building an agent stack from scratch. It also raises the importance of identity, permissions, auditability, approval rules, and data governance. Those concerns are different from choosing a general-purpose chatbot.
OpenAI Dots Pricing and Availability
OpenAI's current announcement says Dots are rolling out to Pro and Business Premium users in eligible markets. Enterprise users, including Edu and Healthcare workspaces, can try the beta when their workspace administrator enables it. The first Dot is included in the Pro or Business Premium plan at no additional charge.
OpenAI says the included Dot is available 24/7 and that plans include an allowance for deeper work, with extended limits during the first month after launch. The company also says that future versions will allow users to add more Dots and scale their output by increasing speed or the total amount of work completed per month.
That means the headline subscription price is not the only factor to examine. Buyers should also check the amount of deeper work included, how connected-app usage is handled, and what additional capacity will cost as the product expands.
OpenAI Dots Safety and Privacy Controls
Persistent agents create a different risk profile from ordinary chat because they can access applications and take actions. OpenAI says Dots have built-in safeguards, configurable permissions, custom rules, activity monitoring, and action review through its auto-review system. The company also says sensitive actions, such as changing a password, remain with the user.
OpenAI says apps used for proactive research are restricted to read-only capabilities when the Dot is working in the background, meaning they cannot send messages, change application content, or control the user's browser or computer through that mode. Users can separately grant permissions for actions they want the Dot to perform.
For business customers, OpenAI says content from Business, Enterprise, and Edu workspaces is not used to improve its models by default. Personal plans have separate controls governing whether Dot conversations and work can be used to improve models.
These controls matter, but they do not remove the need for review. OpenAI itself states that Dots can still make mistakes and recommends reviewing consequential work.
What Early Testers Say About OpenAI Dots
The first public reports provide useful context that the official launch material cannot. BeInCrypto reported that early testers used Dots to sort email, detect scheduling conflicts, and handle administrative work. One tester described permission errors, dropped messages, and missing iMessage support in a pre-launch build and suggested that some users wait before relying on it for important workflows.
These reports should be treated as early product feedback rather than a controlled evaluation. The product was only publicly introduced on September 29, 2026, and availability is still being expanded.
What Makes Dots Different From ChatGPT?
The biggest difference is persistence. ChatGPT is primarily an interactive assistant that responds to prompts, while Dots are designed to continue pursuing a larger objective in the background.
Dots also have a separate cloud computing environment, persistent project context, connected application access, and rules for acting independently or asking for approval. In practice, the product is less about generating one response and more about delegating an ongoing workflow.
That distinction is important for businesses. A company evaluating Dots should compare the product with agent platforms, workflow automation, and managed AI systems, not just with chatbots. ToolJunction's current coverage of enterprise AI agent platforms provides broader context on that market.
Limitations to Consider Before Using OpenAI Dots
Early-stage reliability: Public testing is still new, and early users have reported bugs, permission problems, and dropped messages.
Access is currently limited: Dots are rolling out to selected paid plans and eligible markets rather than being universally available. Enterprise access is currently tied to administrator-enabled beta availability.
Human review still matters: An agent that can take actions creates more risk when it misunderstands a task. OpenAI explicitly recommends reviewing consequential work.
Integration does not equal unrestricted access: The number of connected applications is large, but what a Dot can do depends on the permissions and rules configured for those applications.
Cost will depend on workload: The initial Dot is included with certain plans, but OpenAI says future expansion will involve additional capacity controls. Businesses should evaluate expected workload rather than looking only at the base subscription.
How to Decide if OpenAI Dots Fit Your Workflow
Look for ongoing work, not isolated prompts.
Dots make more sense when a task continues after the initial instruction, such as monitoring customer feedback, updating a proposal, processing recurring administrative work, or maintaining a research workflow.
Map the required permissions.
List the applications the agent needs to access and separate read-only tasks from actions that can change data or communicate externally.
Define approval points.
Sensitive financial, legal, customer-facing, account, or production actions should have clear human review requirements.
Measure the workflow, not just the output.
A useful evaluation should look at how much work the Dot completes, how often it needs intervention, what errors require correction, and how much human time the workflow actually saves.
FAQs About OpenAI Dots
What are OpenAI Dots?
OpenAI Dots are persistent AI agents that can work toward user-defined goals in the background using a cloud computer, browser, and connected applications. They are powered by GPT-6 Astra and are designed to handle ongoing work rather than only answer individual prompts.
How much do OpenAI Dots cost?
The first Dot is included at no extra charge with Pro and Business Premium plans in eligible markets. Enterprise users can access the beta when their workspace administrator enables it. OpenAI says deeper work has plan-based allowances and that future expansion will introduce additional ways to scale output.
What can OpenAI Dots do?
Use cases include software development, research, sales operations, content production, email handling, scheduling-related work, invoicing, and other multi-step workflows. The exact actions available depend on connected applications, permissions, and rules.
Can OpenAI Dots work with Slack and Teams?
Yes. OpenAI says users can message Dots through Slack and Microsoft Teams, while project context can carry across channels.
Are OpenAI Dots always running?
They are designed as always-on agents that can continue working toward goals in the background. OpenAI describes a proactive research mode in which Dots can look for ways to help using connected, read-only application access.
Are OpenAI Dots safe to use for important work?
OpenAI has added permissions, custom rules, monitoring, action review, and approval mechanisms, but the company also says Dots can make mistakes. Important or consequential work should therefore remain subject to human review.
Final Verdict
OpenAI Dots move ChatGPT toward a more persistent agent model. Instead of asking an assistant to complete one task at a time, users can give a Dot an ongoing responsibility and let it work across connected applications while maintaining context.
The strongest early use cases are workflows where information changes over time and the agent can repeatedly collect context, perform work, and bring meaningful decisions back to a person. OpenAI's examples include software fixes, research updates, sales proposals, content production, procurement, customer support, and invoice processing.
At the same time, Dots are still a newly launched product. Early testers have reported useful automation as well as bugs and permission problems, so businesses should validate reliability, permissions, approval flows, and workload limits before handing an agent important responsibility.
The central question is therefore not simply whether OpenAI Dots are more capable than a chatbot. It is whether a persistent agent with access to your real workflows can complete enough of the work reliably, safely, and with appropriate human oversight to justify using it in production.
