OpenAI Launches ‘Dots’, Always-On AI Agents That Work Around the Clock

OpenAI has unveiled “dots”, a new line of always-on artificial intelligence agents for ChatGPT users that keep working on tasks around the clock, the company announced at its DevDay 2026 developer conference in San Francisco on Tuesday.

Chief executive Sam Altman described the agents as “remarkably capable, always-on agents built to handle really anything you can think of”, demonstrating a dot designing websites and slide decks and managing calendars in his keynote, according to Barron’s. Unlike a standard ChatGPT conversation, a dot keeps working independently after the chat ends: it runs on its own cloud computer and browser, remembers context across conversations, learns the user’s preferences over time, and can be given a name and reached by voice, text or messaging apps.

Dots connect to more than 4,000 applications through OpenAI’s plugin system, including calendars and Slack, and stay in sync when a user moves between ChatGPT, Slack and Microsoft Teams, according to Neowin. They are powered by GPT-6 Astra, OpenAI’s newest model.

The feature rolls out from launch day to paying subscribers on ChatGPT Pro and Business Premium plans in eligible markets, with a first dot included at no extra charge, TechCrunch reported; universities, healthcare organisations and businesses can join an admin-gated beta.

The launch comes less than a month after Meta released its own personal agent, Muse, which quickly topped app-store charts — setting up a head-to-head race over the always-on assistant. It also follows OpenAI’s own delay of its GPT-6.1 Astra model a day earlier over failed internal safety tests, a subject Altman did not address on stage, according to the Associated Press.

Sources:

• TechCrunch
• Barron’s
• Neowin
• Associated Press via TechXplore