OpenAI Launches Dots, an Always-On Agentic Assistant Powered by GPT-6 Astra

OpenAI unveiled Dots on Tuesday at its Dev Day event, a new personal agentic assistant powered by GPT-6 Astra. The company describes Dots as "remarkably capable, always-on agents built to handle everything," designed to operate independently of any specific interface or device.

Unlike Codex or ChatGPT, which typically respond to direct prompts within a defined session, Dots are built to pursue user-defined goals continuously in the background with minimal human oversight. According to OpenAI's announcement post, users can "start with your primary dot, give it a name, and make it your own," with plans for "teams of Dots working together" on a user's behalf over time.

Availability and access

Dots launches Tuesday within ChatGPT, limited to Pro and Business Premium subscribers in eligible markets. Users can launch a Dot from either Codex or ChatGPT. OpenAI says it will support messaging Dots through Slack, Teams, and other workplace platforms, with text message support described as coming soon.

Use cases, according to OpenAI

In its announcement, OpenAI outlined several scenarios for how Dots might function:

• A software developer deploys a dedicated Dot to monitor customer feedback and implement bug fixes or requested features as needed.
• A scientist uses a Dot to rerun data analysis and flag unexpected results as new experimental data comes in.

OpenAI also describes plans for "specialist Dots" assigned to specific responsibilities, each provisioned with distinct identities, credentials, and tool access through existing enterprise systems. The company says it is working with Microsoft to integrate Dots with Agent 365, Microsoft's security control framework for agentic systems — though neither company has disclosed technical details or a timeline for that integration.

What's actually new

Much of the underlying functionality — background task execution, tool access, credentialed operation — was already achievable through Codex and comparable agentic harnesses. Dots does not appear to introduce a fundamentally new capability set. Instead, it packages existing agentic infrastructure under a unified product with a persistent identity and a simplified user experience focused on autonomous, ongoing operation rather than session-based prompting.

The product also comes with a distinct visual identity: small, cartoonish, floating "dot" characters, a stylistic choice that echoes Meta's recently released Muse agent. The branding signals a broader industry pattern of companies giving AI agents approachable, character-driven personas rather than purely functional interfaces.

Pricing for Dots beyond existing Pro and Business Premium subscription tiers has not been disclosed, and OpenAI has not specified rate limits, compute costs, or how Dots usage will be metered against existing ChatGPT and Codex quotas.

What this means

Dots is a repackaging effort more than a technical leap: OpenAI is consolidating agentic capabilities already present in Codex into a persistent, branded assistant with its own identity and cross-platform messaging support. The real signal here is strategic — OpenAI is betting that autonomous, always-on agents need a distinct product identity separate from chat interfaces to gain enterprise and consumer traction. The Microsoft Agent 365 integration suggests OpenAI sees credentialing and security controls as the next competitive battleground for agentic AI, not raw model capability. Whether Dots meaningfully outperforms existing Codex-based agentic workflows, or simply wraps them in friendlier branding, remains to be seen once usage data and independent testing become available.