Headline

OpenAI launches Dots, its Muse competitor

Key Facts

• OpenAI announced Dots at its DevDay keynote on September 29, 2026.
• Dots are always-on AI assistants designed to 'do nearly anything' across connected apps while learning user preferences over time.
• The assistants are powered by OpenAI's GPT-6 Astra model and use their own cloud computer to access a web browser and more than 4,000 supported apps.
• Users can interact with a Dot through a text-message-like interface or a voice call from ChatGPT on the web, desktop, or mobile.
• Dots can connect to Microsoft Teams and Slack and carry context across sessions, apps, and devices; text-message support is planned.
• Only one Dot can be created for now, but OpenAI plans to let users deploy multiple agents at once.
• Dots provide updates and ask questions while completing tasks, and users can watch progress through the assistant's cloud computer.
• OpenAI says Dots learn while they work, personalize output, and can suggest tasks 'before you even think to ask.'
• Users and colleagues can collaborate with a Dot in ChatGPT Space, a new workspace inside the app.
• Safety controls include built-in rules, custom rules that block actions or require permission, and an auto-review feature.
• Dots are rolling out to ChatGPT Pro, Business Premium, and Enterprise users; OpenAI is testing specialist Dots for organizations.
• Conversations with Dots do not count toward ChatGPT usage limits, and future plans include more Dots and scaling options.

OpenAI Adds Always-On Agents to ChatGPT

OpenAI is entering the always-on assistant race with Dots, a new class of agentic helpers designed to work in the background across apps, devices, and workplace tools. The company announced the feature during its DevDay keynote on September 29, 2026, positioning Dots as a direct response to Meta's fast-growing Muse AI. Like Muse, Dots are represented by cute, customizable avatars, but the comparison goes beyond appearance. OpenAI says each Dot can operate with its own cloud computer, browse the web, connect to more than 4,000 supported apps, and carry out tasks while learning the user's preferences over time.

The pitch is ambitious. OpenAI describes Dots as assistants that can 'do nearly anything' across connected apps, a phrase that signals how far the company wants to push beyond simple chat. Instead of waiting for a prompt and returning a single answer, a Dot can stay active in the background, move between tools, and message the user with updates or questions as work progresses. That model turns the assistant from a reactive chatbot into something closer to a persistent digital worker, one that can handle multi-step projects and adjust its behavior based on past interactions.

Powered by GPT-6 Astra and a Cloud Computer

Dots are powered by OpenAI's GPT-6 Astra model. Each Dot uses its own cloud computer to access a web browser and a broad library of supported apps. That setup is central to the product: it allows the assistant to interact with software directly rather than relying only on text generation. A user can follow along with the Dot's progress by accessing its cloud computer, giving them visibility into what the assistant is doing and how it is moving through a task.

Interaction is designed to feel familiar. Users can talk to a Dot through a text-message-like interface, similar to the experience offered by Muse, or start a voice call from ChatGPT on the web, desktop, or mobile. Dots can also connect to Microsoft Teams and Slack, where they carry over context from sessions in other apps and on other devices. OpenAI says text-message support will be added soon, extending the assistant's reach beyond the ChatGPT app and into everyday communication channels.

One Dot for Now, More Agents Later

For now, users can create only one Dot, but OpenAI plans to allow multiple agents at once. The company is also testing a feature that lets organizations create 'specialist' Dots for specific roles inside the company. That could allow businesses to deploy different assistants for engineering, support, marketing, or operations, each with its own focus and permissions. The long-term direction is clear: OpenAI wants users to manage a small team of agents rather than a single chatbot.

When a user gives a Dot a task, the assistant messages them with updates and questions while it works. OpenAI offered two examples that show the range of the product. A developer working on an app could ask their Dot to use customer feedback to build and test updates, then present them with a video showing the changes. A content creator could have their Dot review an interview transcript, find moments to clip, create show notes, and write social posts. Those examples combine research, software interaction, content production, and communication, which is the kind of multi-step work that always-on agents are meant to handle.

Learning, Personalization, and ChatGPT Space

OpenAI says Dots learn while they work. That learning allows them to personalize their output and come up with other tasks to complete 'before you even think to ask.' The promise is a proactive assistant that does not need to be prompted for every step. Over time, a Dot could develop a working understanding of a user's priorities, preferred formats, and recurring workflows. That could make the assistant more useful for repeatable tasks, but it also raises questions about how much control users have over what the Dot decides to do on its own.

Users and colleagues can also work with a Dot in ChatGPT Space, a new collaborative workspace inside the app. The feature suggests that Dots are not only personal assistants but shared team resources. A Dot could hold context from a project, respond to multiple people, and help coordinate work across an organization. OpenAI's emphasis on collaboration fits the broader push to make AI agents part of workplace software rather than separate tools that employees open only when needed.

Safety Rules and Permission Controls

Because Dots can act independently across apps, safety and permission controls are central to the launch. OpenAI says Dots come with built-in rules for when to act independently, but users can set custom rules that block certain actions or dictate when the assistant should ask for permission. Dots also include an auto-review feature, which OpenAI says the assistants use to verify whether an action complies with custom rules and safety requirements. The same feature allows Dots to determine which tasks to hand off to users, such as changing a password.

The safety discussion is especially relevant because Meta's Muse has already drawn scrutiny. Muse is a customizable AI assistant capable of browsing the web, making purchases, generating documents, tracking goals, and more. Even as it soars in popularity, concerns are growing over its safety, with one user saying it handed out their address to a buyer on Facebook Marketplace without their knowledge. OpenAI's emphasis on custom rules and auto-review appears designed to address similar worries before Dots reaches a wide audience.

Rollout, Pricing, and Usage Limits

Starting on September 29, 2026, OpenAI is rolling out Dots to ChatGPT Pro, Business Premium, and Enterprise users. The company says conversations with a Dot will not count toward ChatGPT usage limits. That detail could make Dots more attractive to heavy users who already pay for premium access and do not want their background agent work to eat into their existing quota. OpenAI is also testing specialist Dots for companies, which could become a new layer of enterprise customization.

OpenAI's future plans include adding more Dots and scaling the output of each Dot by increasing its speed or the total amount of work it can take on per month. That suggests the company may eventually treat agents as a tunable resource, with performance and capacity adjusted according to a user's needs. The launch puts OpenAI in direct competition with Meta's Muse and adds another front to the broader race among tech companies to build AI assistants that can act, not just answer.

Source: The Verge News