Anthropic's new agent does not wait for a prompt — it watches your work apps, sends a morning brief, and asks before it acts, in a direct challenge to Meta Muse and OpenAI Dots.
October 22, 2026 — Anthropic has launched Claude Duet, a proactive AI agent that reaches out before you ask. Announced on October 21, 2026, Duet is built on Claude Opus 5.5 and is opening as a research preview for Pro and Max subscribers in the United States, through a waitlist.
Image: A proactive AI assistant sending a morning brief from connected work apps on phone and laptop. Credit: AI-generated illustration / Red Eye Eagle News.
Editor's Highlight: The important shift is not a smarter chatbot. It is timing. Duet is designed to start work in the background, flag what needs attention, and only interrupt its user when a decision is needed — while still asking for approval before it sends, pays, signs or deletes anything.
Table of Contents
- Key Facts
- What Claude Duet Does Differently
- How Duet Works Through the Day
- Who Can Try It and What It Costs
- Safety, Approvals and Hard Limits
- A Crowded Race for Proactive Agents
- What to Watch Next
- Bottom Line
- FAQ
- Sources
Key Facts
| Product | Claude Duet, a standalone proactive AI agent from Anthropic |
| Announced | October 21, 2026, as a research preview |
| Model | Built around Claude Opus 5.5 |
| Who can access it | Pro and Max subscribers in the United States, via a waitlist |
| Where it runs | Web and standalone apps; works in the cloud on its own virtual machine with its own browser |
| Core routine | Morning brief, then checks connected apps every few hours and alerts when something needs attention |
| Connections | More than 6,000 connectors and plug-ins in the Claude Marketplace, according to Anthropic |
| Cost in preview | No extra cost beyond the plan; background work counts toward existing plan limits |
What Claude Duet Does Differently
Most assistants wait for a prompt. Claude Duet reverses that order, according to Anthropic. It connects to apps, suggests what needs attention, and completes tasks in the cloud.
Anthropic says it built Duet for professionals and small business owners with too much to track.
Newsroom Focus: Duet is a standalone product, not just a new mode inside regular Claude. It carries over memory, skills and connected apps from a Claude account, so setup is limited, and interactions happen in one ongoing thread.
How Duet Works Through the Day
Each morning, Duet sends a short brief on important items from connected apps. It then checks those apps every few hours.
When something needs attention, it sends an alert. When nothing does, it stays quiet.
Checking constantly would use plan limits much faster, which is why periodic checks matter.
| Stage | What Duet does |
| Morning | Sends a brief on the most important items across connected apps |
| Through the day | Checks apps every few hours and messages only when something needs attention |
| Task work | Runs on its own cloud virtual machine and browser, even when a phone or laptop is off |
| Completion | Notifies the user when work is done or needs review |
Tasks range from scheduling and drafting replies to building dashboards, financial models and web apps.
Who Can Try It and What It Costs
Access is narrow at launch. Duet is in research preview, granted via a waitlist, and limited to Pro and Max subscribers in the United States.
It can be used via the web and standalone apps, with iOS and Android access described in launch coverage.
There is no separate Duet price in the preview. Instead, Duet uses existing Claude plan limits, and its background checks count toward them.
Practical Takeaway: That pricing detail is a real limit. An agent that checks apps all day and runs long tasks will consume the same allowance a subscriber already uses for regular Claude work.
Safety, Approvals and Hard Limits
An agent that reads email and prepares payments needs clear brakes. Anthropic's design, as reported, puts approval first for sensitive steps.
Duet will get user approval before sending a message, making a payment, signing a document, or deleting a file. Some actions are off-limits even with approval, including trading securities and moving money between accounts, according to the company.
- It does not see passwords or logins when given access to accounts, according to Anthropic.
- It has a dedicated classifier and safety system prompt that checks each message.
- That system is designed to guard against sensitive information leaks.
- Purchases can be made through Stripe's Link wallet or Shop Pay, with approval required.
Analysis Note: These are Anthropic's claims, not a guarantee. Coverage also notes Duet can misread a conversation, misjudge urgency, or miss something between checks — so users must review approvals carefully.
A Crowded Race for Proactive Agents
Duet does not arrive alone. It is Anthropic's answer to Meta's Muse agent and OpenAI's Dots agent, both described as proactive agents that reach out before being asked.
| Product | Company | Positioning in reports |
| Claude Duet | Anthropic | Work-focused proactive agent for professionals and small businesses |
| Dots | OpenAI | Proactive agent with a business slant and professional tool integrations |
| Muse | Meta | Personal AI agent, with a version aimed at small businesses |
TechCrunch framed Duet as Anthropic's bet on agents. Competition will now turn on reliability, integrations and trust, not just model quality.
TechCrunch also reported that Duet is not positioned for personal errands or booking tickets, and users cannot watch its browser use as it happens.
What to Watch Next
Watch three things: how quickly the waitlist opens, whether background work drains plan limits too fast, and whether Duet catches genuinely important items without creating extra supervision work.
Bottom Line
Claude Duet marks Anthropic's move from a chatbot you visit to an agent that visits you. Built on Opus 5.5, running in the cloud, and bundled into existing Pro and Max plans during its US research preview, it offers a clear idea with clear limits.
What to watch is simple: whether users let an agent monitor their apps every few hours — and whether its morning brief earns that trust.
FAQ
What is Claude Duet?
It is Anthropic's standalone proactive agent. It connects to apps, sends a morning brief, checks for items needing attention, and completes tasks in the cloud.
Who can use Claude Duet now?
Only US Pro and Max subscribers can join the waitlist. Access is granted in stages, so joining does not guarantee immediate entry.
Does Claude Duet cost extra?
No extra cost was announced for the preview. Duet is included with eligible Pro and Max plans, but its background checks and tasks count toward the plan's existing usage limits.
Can Duet send messages or make payments on its own?
According to Anthropic, no. It asks for approval before sending, paying, signing or deleting. Trading securities and moving money between accounts are off-limits even with approval.
How is Duet different from regular Claude?
Regular Claude responds when prompted. Duet initiates contact, works in the background on its own virtual machine, and carries over memory, skills and connected apps into one ongoing thread.
Sources
- MacRumors: https://www.macrumors.com/2026/10/21/anthropic-claude-duet/
- TechCrunch: https://techcrunch.com/2026/10/22/claude-duet-app-is-anthropics-bet-on-agents/
- iTechPost: https://www.itechpost.com/articles/237623/20261022/anthropics-claude-duet-ai-agent-launch-signals-new-battle-openai-meta.htm
- TechBooky: https://www.techbooky.com/anthropic-launches-claude-duet-a-proactive-ai-agent-for-work/
- Image: AI-generated illustration / Red Eye Eagle News.
Join the conversation
Comments
Post a Comment