OpenAI Dots: GPT-6 Astra Agents with Persistent Cloud Environments

Seed story: "Introducing dots - OpenAI" · search original Written from facts verified across 3 report(s) — original explainer, not a copy or translation. Sources at the end.

OpenAI has introduced Dots, a new class of always-on agentic assistants powered by the GPT-6 Astra model that operate with their own persistent cloud computer and browser. By shifting autonomous coding and data analysis tasks into dedicated, stateful environments accessible via ChatGPT, Slack, and Microsoft Teams, Dots aim to reduce the friction of managing context for developers working on long-running, multi-step goals.

Dots Launch and Core Architecture

OpenAI officially unveiled Dots at its DevDay event on September 29, 2026, marking a significant shift toward autonomous personal assistants. These agents are powered by the GPT-6 Astra model and are engineered to operate continuously in the background, pursuing user-defined goals with minimal oversight.

Unlike traditional chat interfaces, each dot is provisioned with a dedicated cloud computer and browser. This persistent infrastructure allows the agent to maintain state and execute complex workflows independently. Key architectural features include:

Always-on operation for background task execution
Dedicated cloud computing resources
Integrated browser access for web-based actions

For developers, this architecture implies a new paradigm for tooling. Instead of invoking models via API calls for discrete tasks, teams can now deploy persistent agents that retain context and environment state. This setup supports long-running processes, such as continuous data analysis or software development, where the agent actively manages its own workspace rather than waiting for explicit prompts.

Persistent State and Context Management

Traditional LLM interactions are inherently stateless, forcing developers to manually reconstruct context for every new session. Dots eliminate this friction by operating within a dedicated cloud computer. This persistent environment acts as a continuous workspace, allowing the GPT-6 Astra model to maintain long-term memory and a stable file system across interactions.

Instead of resetting after each query, the agent retains access to previous work artifacts and project files. This capability is critical for complex, multi-step tasks that span days or weeks.

Persistent File System: Agents retain access to codebases and documents without re-uploading.
Long-Term Context: Background processes maintain awareness of ongoing goals and historical decisions.
Continuous Execution: Workflows proceed independently, reducing the need for constant human prompting.

For developers, this shifts the paradigm from discrete chat sessions to a persistent, autonomous workspace where state is managed by the infrastructure rather than the user.

Integration with Microsoft Agent 365

Enterprise Security and Identity

Integrating Dots with Microsoft Agent 365 addresses critical enterprise concerns by leveraging existing security controls and identity management infrastructure. Rather than building a siloed authentication system, OpenAI is embedding these agents within the broader Microsoft ecosystem. This approach ensures that enterprise compliance controls are applied consistently, reducing the risk of unauthorized access or data leakage in sensitive business environments.

For developers and IT administrators, this integration simplifies deployment and governance. Key implications include:

Unified Identity: Agents inherit user permissions, preventing privilege escalation.
Compliance: Existing audit logs and data residency policies apply to agent actions.
Security: Standard enterprise security controls manage agent behavior.

This alignment allows organizations to adopt autonomous agents without compromising their established security posture.

Multi-Channel Interaction and Plugin Ecosystem

Dots break the silo of single-interface assistants by meeting developers where they already work. You can interact with your agent directly through ChatGPT, Slack, or Microsoft Teams, ensuring that autonomous tasks remain accessible within your existing communication workflows. While text message support is planned for the future, these initial channels allow for immediate integration into daily operations without requiring new, dedicated applications.

The true power of this multi-channel approach lies in the underlying plugin ecosystem. By connecting to over 4,000 apps, Dots extend their capabilities far beyond simple chat responses. This extensive connectivity enables agents to perform complex, real-world tasks such as:

Conducting deep research across multiple platforms
Executing advanced data analysis
Assisting with software development workflows

For developers, this means your agent can pull context from various tools simultaneously, acting as a unified hub for information retrieval and execution rather than a standalone chatbot.

Implications for Autonomous Coding Workflows

From Command to Oversight

The shift from direct command to goal definition represents a fundamental change in how developers interact with software. By operating independently in the background, Dots allow engineers to step away from the keyboard and focus on high-level objectives rather than line-by-line execution. This autonomy means that the developer's primary responsibility transitions from writing code to defining the desired outcome and setting clear success criteria.

This new workflow requires a different set of skills:

Precise goal articulation to guide autonomous agents
Rigorous result verification to ensure quality
Strategic oversight of background processes

Consequently, the role evolves from a hands-on coder to a supervisor who validates the agent’s work. This approach can accelerate shipping cycles by offloading routine tasks, but it demands that developers maintain strong verification practices to catch errors that might slip through in a low-oversight environment.

Rollout Scope and Developer Access

OpenAI is gating the initial rollout of Dots to Pro and Business Premium subscribers in eligible markets. This phased approach allows the company to test the stability of always-on agents before broadening access. Developers on these tiers gain immediate entry to a persistent cloud environment, effectively turning their subscription into a dedicated, autonomous workspace.

The roadmap outlines a clear path for wider adoption and deeper integration:

Expansion to Enterprise plans is planned to bring these capabilities to larger organizations.
Text message support is scheduled for future releases, extending interaction beyond current channels.
Integration with Microsoft Agent 365 will leverage existing security controls and identity management.

For teams, this means the tooling will eventually align with corporate identity standards. As Enterprise access arrives, developers can expect their autonomous coding agents to operate within established security perimeters, reducing the friction of deploying persistent, background-driven workflows in regulated environments.

FAQ

What model powers OpenAI Dots and how do they operate?

OpenAI Dots are powered by the GPT-6 Astra model and function as always-on agents. They operate with their own persistent cloud computer and browser to work independently in the background on user-defined goals.

Who can access OpenAI Dots and how can users interact with them?

The service is initially rolling out to Pro and Business Premium users in eligible markets, with future plans for Enterprise plans. Users can interact with their dot through ChatGPT, Slack, and Microsoft Teams, while text message support is planned for the future.

How do OpenAI Dots integrate with enterprise security and external applications?

OpenAI is integrating specialist Dots with Microsoft Agent 365 to leverage existing security controls and identity management. Additionally, Dots can connect to over 4,000 apps via an ecosystem of plugins to perform tasks like research and software development.