OpenAI has launched Dots, a new personal AI agent system powered by GPT-6 Astra, designed to handle tasks continuously with minimal user oversight.
The company introduced Dots during its Dev Day event, describing them as always-on agents that can work independently toward user-defined goals rather than being limited to a specific app or interface.
AI Agents Designed to Work in the Background
OpenAI said users can create a primary Dot, name it, and customise how it works. The company’s long-term vision includes multiple Dots working together to manage different responsibilities for users.
Dots can be launched through ChatGPT or Codex and are initially available for Pro and Business Premium users in eligible markets.
OpenAI highlighted several potential uses for Dots, including software developers using agents to monitor customer feedback and implement improvements, or researchers using agents to rerun analyses and investigate new experimental results.
Enterprise Workflows and Specialist Agents
Users will be able to interact with Dots through workplace platforms such as Slack and Teams, with text messaging support planned for the future.
OpenAI also envisions specialist Dots that handle specific responsibilities. These agents can be configured with dedicated identities, credentials, and tools through existing enterprise systems.
The company is working with Microsoft to integrate Dots with Agent 365 security controls, helping organisations manage AI agents within business environments.
OpenAI Gives AI Agents a New Identity
Many of the capabilities behind Dots were already possible through Codex and other AI agent frameworks, but OpenAI is packaging them as a simpler experience focused on independent actions.
The launch also introduces a new visual identity for OpenAI’s agents, with Dots represented as small, animated characters designed to make AI assistants feel more approachable and easier for users to interact with.
