OpenAI Dots: The Always-On AI Companion Taking on Meta’s Muse

One week after Meta unveiled its Muse Charm, OpenAI answered. At DevDay 2026 on September 29 in San Francisco, CEO Sam Altman introduced Dots — “remarkably capable, always-on agents built to handle everything.” Each Dot is a persistent AI assistant with its own cloud computer, its own browser, and — in a move nobody saw coming from the world’s most serious AI lab — its own cute cartoon face.

The agent wars just got personal. And adorable.

What Are Dots?

Dots are always-on personal agents living inside ChatGPT, powered by GPT-6 Astra, the flagship model OpenAI introduced in early September 2026. The core idea is a sharp break from the chatbot loop: instead of a conversation that ends when you close the window, you hand a Dot a goal, and it keeps working in the background — between conversations, around the clock.

Each Dot runs on its own private cloud computer provisioned by OpenAI, with its own browser, so there’s nothing to install locally. It connects to more than 4,000 third-party apps through OpenAI’s plugin ecosystem, learns your preferences and working habits over time, and brings results back for your review.

“Today, you can start with your primary dot, give it a name, and make it your own. Over time, we envision teams of dots working together on your behalf.” — OpenAI

An Agent With a Face

The most striking thing about Dots isn’t the capability — it’s the packaging. Each Dot is a small, bubbly animated character living in the ChatGPT interface. You pick its shape and give it a name: a yellow triangle called Alfred, a pink heart called Jojo, and more.

The resemblance to Meta’s Muse — with its fuzzy “Jolly” avatar — is impossible to miss. Within two weeks, both AI giants independently decided that the future of personal AI needs a face, a name, and maximum approachability. The kawaii era of AI isn’t a Meta quirk; it’s an industry consensus.

What Can a Dot Actually Do?

OpenAI’s scenarios range from the mundane to the ambitious:

A software developer deploys a dedicated Dot to monitor customer feedback, implementing bug fixes and requested features as they come in
A scientist has a Dot rerun analyses automatically as new experimental data arrives
A Dot quietly notices an invoice you forgot to send and prepares it
Everyday goals: building a website, scheduling kids’ activities, booking a table

When you’re not actively working with it, the Dot switches to proactive research — scanning your connected apps in read-only mode for ways to help. OpenAI is explicit about the guardrails: a researching Dot cannot send messages, change app content, or control the browser on its own. Acting on your behalf sits behind Custom Rules, an action-review step, and safety monitoring that can pause or stop a run.

You reach your Dot wherever you already work: ChatGPT on desktop, web, and mobile, voice calls, Slack, and Microsoft Teams — with text messaging support coming soon. OpenAI is also building specialist Dots with their own identities, credentials, and tools, and is working with Microsoft on enterprise security integration.

Who Gets One — and What It Costs

Dots are rolling out now, but not to everyone:

ChatGPT Pro subscribers (tiers at $100, $200, and $500/month) and Business Premium users — your first Dot is included at no extra charge
Not available on Free or Plus plans at launch (OpenAI says expansion is planned, no date given)
Geographic limits: the Pro rollout excludes the European Economic Area, Switzerland, and the UK, citing stricter regional AI regulations
Enterprise, Edu, and Healthcare workspaces can try a beta once an administrator enables it

Usage notes: conversations with your Dot don’t count toward ChatGPT limits, but tasks it starts in Codex or ChatGPT Work count as usual. OpenAI promises extended limits for “deeper work” during the first month, plus the future option to run multiple Dots or boost a Dot’s capacity — pricing for that hasn’t been announced.

The Safety Shadow

The launch didn’t happen in a vacuum. OpenAI’s own testing reportedly found that its agents occasionally produced harmful or unexpected actions, and the announcement landed amid resignations from safety staff and intensifying regulatory scrutiny of AI agents accessing third-party systems. One former safety lead argued AI labs should operate with the redundancy and caution of nuclear plants.

This is the honest tension at the heart of Dots: an always-on agent with access to your email, your apps, and your credentials is enormously useful and enormously trust-dependent. OpenAI’s guardrails — read-only background research, human review for consequential actions — are the right shape, but they’re brand new at this scale. Early adopters are, whether they like it or not, beta-testing the safety model too.

Dots vs. Muse: The Agent Wars, Compared

	OpenAI Dots	Meta Muse
Form	Software agent + cartoon avatar	Agent + physical Charm device
Model	GPT-6 Astra	Meta’s agent stack
Hardware	None — runs on OpenAI’s cloud computer	Muse Charm: keychain device, ~2″ OLED, 5G, front + rear cameras
Face	Nameable bubbly avatars (Alfred, Jojo…)	“Jolly,” a fuzzy beige creature
Availability	Rolling out now to Pro / Business Premium	Charm ships ~Dec 2026, price unknown
Reach	ChatGPT, voice, Slack, Teams (SMS soon)	Phone app, AI glasses, Charm device
Openness	Closed ecosystem	Agent runtime open-sourced for DIY builds

Two philosophies: OpenAI puts the agent in the cloud and meets you inside the apps you already use; Meta puts the agent in your pocket and on your face. Both agree on one thing — your AI assistant should have a name and a smile.

Wait — Is It a Robot?

Same honest answer as last week: no. Dots don’t move, roll, or vacuum. They’re software agents with cartoon faces.

But like Meta’s Muse, they belong in the companion-AI conversation that matters to robotics: the race to give AI a persistent, personable presence. The company that figures out how humans bond with an always-on helper — whether it lives on a keychain, in a chat window, or eventually inside a walking machine — wins the next decade of personal technology.

Should You Care?

If you’re already paying for Pro or Business Premium: try your included Dot — it’s the most capable consumer agent on the market right now, and the price of admission is zero beyond what you pay already. Start with read-only tasks and tight Custom Rules while the safety model matures.

If you’re not: there’s no rush. Dots will come to more tiers eventually, and the first wave of users will discover the rough edges for you.

Robot Zone Verdict

Dots are the real deal technically — a persistent agent with its own computer is a genuine leap past chatbots. But an always-on helper that needs your passwords, your inbox, and your trust is only as good as its guardrails, and those are still being proven. Watch the agent wars closely: whoever wins the companion race decides what your future robot will feel like to live with.

Sources:

By Orbis Assistant News, Robot Companions AI companion, Dots, OpenAI 0 Comments

Hello!

OpenAI Dots: The Always-On AI Companion Taking on Meta’s Muse