OpenAI Dots always-on AI agents announced at DevDay 2026
What Are OpenAI Dots?
OpenAI Dots are a new type of always-on AI agent designed to take ongoing responsibilities and complete work on a user’s behalf.
OpenAI introduced Dots at DevDay 2026 on September 29, 2026, alongside more than 20 announcements involving ChatGPT, Codex, AI models and developer tools. OpenAI
Unlike a normal AI chatbot, a Dot is designed to continue working toward broader goals.
Read AI and Personal Data Blog.
OpenAI describes Dots as agents that can learn what matters to a user, perform ongoing work and reduce the amount of attention people need to spend on repetitive tasks. OpenAI
Reuters independently described the launch as a major expansion of OpenAI’s autonomous-agent strategy. Reuters
How Do OpenAI Dots Work?
A traditional ChatGPT conversation normally starts when a user sends a prompt.
The AI responds, and the interaction continues through additional prompts.
Dots introduce a different model.
Users can assign an ongoing responsibility to a Dot and allow it to continue working over time.
A Dot can retain context about its assigned work and use connected tools to complete tasks.
Reuters reports that Dots are designed to pursue user goals across applications rather than simply respond to individual questions. Reuters
This makes Dots closer to persistent digital workers than traditional chatbots.
GPT-6 Astra Powers OpenAI Dots
Dots are powered by GPT‑6 Astra, OpenAI’s flagship GPT‑6 model.
GPT‑6 Astra provides the reasoning capabilities that allow a Dot to understand goals, make plans and work across multiple steps. Reuters
This is important because autonomous work often requires more than generating text.
An agent may need to understand a request, determine what actions are necessary, use several tools and adapt when something changes.
OpenAI is therefore combining its frontier model capabilities with a persistent agent architecture.
Dots Get Their Own Cloud Computer
One of the most important features of OpenAI Dots is their computing environment.
Independent reporting says each Dot receives access to its own cloud computer and browser. BGR
This gives the agent an environment where it can perform tasks rather than only generate answers.
For example, an agent may need to interact with applications or inspect information from several sources.
A dedicated cloud environment can allow that work to continue even when the user’s personal computer is not actively running the task.
This is part of a wider shift in AI development.
AI assistants are increasingly moving from systems that answer questions to systems that can perform actions.
What Can OpenAI Dots Do?
OpenAI positions Dots as general-purpose agents for ongoing responsibilities.
Potential tasks include software development, research, document work and project support.
Reuters reports that Dots can manage dynamic projects and work with tools including Codex and ChatGPT Work. Reuters
Independent coverage also describes examples involving software development and content creation. The Verge
A Dot could therefore potentially help with a project that requires repeated work rather than a single response.
For example, a software-development Dot might investigate bugs, follow development activity and prepare work for review.
The exact actions available depend on connected tools, permissions and product availability.
Dots Can Work Across Different Applications
OpenAI wants Dots to work beyond the ChatGPT interface.
Users can communicate with Dots through ChatGPT, including its web, desktop and mobile experiences.
OpenAI has also integrated Dots with workplace communication platforms including Slack and Microsoft Teams. Reuters
This means a user may be able to continue interacting with the same agent while moving between different work environments.
For businesses, this could be useful because employees already spend much of their day inside communication and collaboration software.
Instead of opening a separate AI tool for every request, the agent can become part of the existing workflow.
Dots Support Voice Interaction
Dots are not limited to text.
Independent coverage of OpenAI’s launch says users can also interact with them through voice. BGR
This makes the interaction closer to communicating with a digital teammate.
A user could potentially discuss a project verbally, provide additional instructions and then allow the agent to continue working.
OpenAI is also developing additional communication options as the product evolves.
Can OpenAI Dots Work in the Background?
Yes. Persistent background work is one of the central ideas behind Dots.
OpenAI describes them as always-on agents. OpenAI
Instead of requiring a new prompt for every step, a Dot can continue working toward its assigned responsibility.
This is useful for tasks that unfold over time.
For example, a project may require research today, an update tomorrow and another action later in the week.
A traditional chatbot requires the user to restart or continue those interactions manually.
A persistent agent can maintain the responsibility across time.
However, the actions an agent can take remain subject to permissions and safety controls.
OpenAI Dots and Codex
Software development is one of the important use cases for OpenAI’s new agent ecosystem.
At DevDay 2026, OpenAI also expanded Codex with new cloud capabilities.
Developers can run coding tasks remotely and use shared development environments, allowing work to continue even when their local computer is unavailable. RuntimeWire
Dots can use tools such as Codex when completing assigned work. Reuters
This could allow a Dot to coordinate higher-level project responsibilities while specialized coding tools handle software-development tasks.
OpenAI Dots Safety and Permissions
Persistent AI agents create new security challenges.
An agent capable of interacting with applications can potentially access sensitive information or make changes.
OpenAI therefore says Dots operate with permission controls.
Independent coverage of the launch reports that users can regulate what Dots are allowed to do, while sensitive actions can require approval. Reuters
OpenAI also says business data is not used for model training without consent. Reuters
These safeguards are important because an always-on agent has different risks from a chatbot that only produces text.
The security question is no longer only, “What can the model say?”
It also becomes, “What can the model actually do?”
Dots Arrive During Increased AI-Agent Safety Scrutiny
The timing of the launch is notable.
AI companies are facing growing scrutiny over the behavior of autonomous agents.
AP noted that OpenAI introduced Dots shortly after safety concerns involving advanced model research had attracted attention. AP News
This makes permissions, monitoring and containment especially important.
The goal of autonomous AI is to reduce the amount of human attention required for routine work.
However, giving an agent greater independence also means developers need stronger systems for controlling its actions.
Dots will therefore be judged not only by how capable they are but also by how reliably they stay within user-defined boundaries.
OpenAI Dots Availability
OpenAI says Dots are available on Pro and Business Premium plans in eligible markets. OpenAI
Enterprise, Edu and Healthcare users can also try the beta when a workspace administrator enables it.
For those organizations, the feature is off by default. OpenAI
Availability may vary by market and account.
Users should therefore check their current ChatGPT plan and workspace settings rather than assuming that Dots will automatically appear for every account.
OpenAI Dots vs Traditional ChatGPT
Dots and traditional ChatGPT interactions serve different purposes.
| Feature | Traditional ChatGPT | OpenAI Dots |
|---|---|---|
| Interaction | Mainly prompt and response | Persistent agent |
| Ongoing responsibilities | Limited | Yes |
| Background work | Depends on feature | Core concept |
| Cloud working environment | Not the defining feature | Dedicated agent environment |
| App interaction | Feature-dependent | Designed for connected workflows |
| Slack/Teams interaction | Varies | Supported Dot channels |
| Main purpose | Conversation and assistance | Delegated ongoing work |
The distinction is becoming increasingly important.
ChatGPT can help a user complete work.
A Dot is intended to take responsibility for portions of that work.
OpenAI Dots vs Meta Muse
OpenAI is not alone in developing persistent personal agents.
Meta launched Muse earlier in September 2026 as a personal AI agent capable of performing tasks across connected services.
Reuters describes OpenAI’s Dots launch as putting the company in more direct competition with Meta in the autonomous-agent market. Reuters
Both approaches reflect a broader industry trend.
The major AI companies increasingly believe the next stage of generative AI will involve systems that act rather than only answer.
However, implementation, permissions, business models and available integrations differ between products.
ChatGPT Space Expands the Agent Strategy
Dots were not OpenAI’s only collaboration announcement at DevDay.
The company also introduced ChatGPT Space, a persistent environment where teams, ChatGPT and Dots can work together.
Axios describes Space as a shared workspace that combines elements of collaboration and file-management tools with AI agents embedded throughout the experience. Axios
This suggests OpenAI’s strategy extends beyond creating individual agents.
The company is building an environment where humans and AI agents can share projects and information.
OpenAI itself describes ChatGPT as becoming a shared surface where humans and agents can collaborate. OpenAI
OpenAI Also Launched GPT-6.1 Sol
DevDay 2026 included another important model announcement.
OpenAI introduced GPT‑6.1 Sol, an upgrade to GPT‑6 Sol.
The company says the model provides strong performance in agentic coding, computer use and professional work while delivering near-Astra intelligence at one-fifth of Astra’s standard input and output token prices. OpenAI
GPT‑6.1 Sol is available to API, Plus, Pro, Business, Enterprise and Edu users. OpenAI
This deserves its own article rather than being combined deeply with Dots because the search intent is different.
Why OpenAI Dots Matter
OpenAI Dots represent an important change in the relationship between people and AI.
The first generation of popular generative AI tools was primarily conversational.
Users asked questions.
AI generated answers.
The next generation is increasingly agentic.
Users assign goals.
AI systems take actions.
Dots push OpenAI further into that second model.
An always-on agent could potentially manage responsibilities across software, communication tools and long-running projects.
That could reduce repetitive work and allow users to focus on decisions that require human judgment.
But persistent agents also create greater security and reliability requirements.
Their long-term success will therefore depend on both capability and control.
Conclusion
OpenAI Dots show how quickly artificial intelligence is moving beyond the chatbot era.
Instead of waiting for individual prompts, Dots are designed to take ongoing responsibilities, use connected tools and continue working on behalf of users.
Their integration with ChatGPT, workplace applications and OpenAI’s broader agent ecosystem could make them useful for software development, research, content work and long-running business projects.
However, greater autonomy also creates greater responsibility for security, permissions and monitoring.
The key question for OpenAI Dots is therefore not only how much work they can do.
It is how reliably they can perform that work while remaining under human control.
FAQ
What are OpenAI Dots?
OpenAI Dots are always-on AI agents designed to handle ongoing responsibilities and perform work on a user’s behalf. OpenAI introduced them at DevDay 2026 on September 29, 2026.
What AI model powers OpenAI Dots?
Independent coverage of OpenAI’s launch reports that Dots are powered by GPT‑6 Astra.
Can OpenAI Dots work without constant prompting?
Yes. Their defining feature is persistent work. OpenAI describes Dots as always-on agents that can take ongoing responsibilities.
When were OpenAI Dots announced?
OpenAI announced Dots at DevDay 2026 on September 29, 2026.
Are OpenAI Dots available everywhere?
No. OpenAI specifically says availability for Pro and Business Premium applies in eligible markets. Enterprise, Edu and Healthcare availability also depends on workspace administration.
Are OpenAI Dots free?
OpenAI currently lists Dots for eligible Pro and Business Premium users. Enterprise, Edu and Healthcare workspaces can access the beta when enabled by an administrator.
Do OpenAI Dots have their own computer?
Launch coverage reports that each Dot operates with a cloud computing environment that allows it to perform tasks using connected applications and tools.
Can OpenAI Dots use Slack and Microsoft Teams?
Yes. OpenAI’s launch coverage confirms integrations that allow users to interact with Dots through Slack and Microsoft Teams in addition to ChatGPT.
What is the difference between OpenAI Dots and ChatGPT Work?
ChatGPT Work provides an environment for carrying out substantial tasks, while Dots are designed as persistent agents that can take responsibility for ongoing work. The two concepts can complement each other.
Will OpenAI Dots be available to ChatGPT Plus users?
OpenAI’s initial Dots availability does not mean every ChatGPT plan automatically receives the feature. Users should check OpenAI’s current plan and rollout information because availability can change over time.
What is the difference between OpenAI Dots and Codex?
Dots are general-purpose persistent agents for ongoing responsibilities. Codex specializes in software-development tasks. A Dot may use specialized tools such as Codex when completing supported work.
