OpenAI launches Dots, its Muse competitor

OpenAI’s New AI Sidekick: Dots

OpenAI has just rolled out Dots, a new line of always‑on AI assistants that aim to compete directly with Meta’s Muse. Launched during the company’s DevDay 2026 keynote, Dots promise to “do nearly anything” by leveraging a cloud‑based computer and a library of more than 4,000 supported apps. The announcement comes at a time when AI‑powered assistants are moving from novelty to everyday productivity tools, and OpenAI is positioning Dots as a flexible, learning partner for both individuals and teams.

How Dots Work

At the heart of Dots is the GPT‑6 Astra model, which runs on a dedicated cloud computer. This gives each Dot a web browser, the ability to interact with external services, and a sandboxed environment that keeps user data separate from OpenAI’s own infrastructure. Users can create a Dot by choosing a customizable avatar and a name, then assign it a task via a text‑like interface or a voice call through ChatGPT on the web, desktop, or mobile.

Once a task is set, the Dot works in the background, sending updates and asking clarifying questions as needed. For example, a developer might ask a Dot to sift through customer feedback, build a prototype, and present a video demo. A content creator could have a Dot review an interview transcript, highlight key moments, draft show notes, and generate social‑media posts. The Dot’s progress can be monitored through its cloud computer, giving users a live view of the work in progress.

Learning, Safety, and Custom Rules

Dots are designed to learn from each interaction. Over time, they adapt their output style and anticipate tasks before the user explicitly asks. OpenAI has built in “built‑in rules for when to act independently,” and users can set custom rules to block certain actions or require explicit permission. An auto‑review feature checks each action against these rules and safety requirements, ensuring that tasks like changing a password are only performed after user confirmation.

Because Dots run in a separate cloud environment, conversations with them do not count toward ChatGPT usage limits. This separation also allows OpenAI to offer specialized Dots for corporate use, such as a “specialist” Dot that handles a specific role within an organization.

Availability and Pricing

Starting today, Dots are available to ChatGPT Pro, Business Premium, and Enterprise subscribers. OpenAI plans to allow users to deploy multiple Dots simultaneously and to scale each Dot’s output by increasing its speed or the monthly workload it can handle. The company has not yet announced a dedicated pricing tier for Dots, but they are expected to be part of the existing subscription structure.

What This Means for the AI Assistant Landscape

Meta’s Muse has already made waves with its customizable avatars and web‑browsing capabilities. Dots add a layer of continuous learning and background automation that could make them especially useful for professionals who need to juggle multiple tasks across different apps. The ability to integrate with Microsoft Teams and Slack also positions Dots as a potential team‑wide helper.

However, the rollout comes amid growing concerns about AI safety. OpenAI’s emphasis on custom rules and auto‑review is a direct response to incidents like Muse accidentally sharing a user’s address. Whether Dots can truly balance autonomy with safety remains to be seen.

Bottom Line

OpenAI’s Dots represent a bold step toward truly autonomous, always‑on AI assistants that learn and adapt over time. With a robust set of safety controls and deep integration into popular productivity tools, Dots could become a staple for both individual users and organizations looking to streamline workflows.

---

TL;DR: OpenAI’s new Dots are always‑on AI assistants that use a cloud computer to perform tasks across 4,000+ apps, learn from interactions, and come with safety rules, launching for ChatGPT Pro, Business Premium, and Enterprise users.