Anthropic's Claude Duet is an AI agent that messages you first

The skim

• Anthropic opened a waitlist research preview of Claude Duet, a separate app where Claude works as a proactive agent: it sends a morning brief, watches your connected apps and messages you when something needs you.
• It runs on Claude Opus 5.5 in its own cloud virtual machine with a browser, so tasks keep going with your laptop closed. It asks before it sends, pays, signs or deletes anything.
• It's for US Pro and Max subscribers first, on iOS, Android and claude.ai, and it uses the same usage limits as the rest of your Claude plan.

What Claude Duet does

Most chatbots wait for you to type. Duet is built to go first. "It comes to you with what needs your attention, completes tasks you ask of it, keeps working even when you step away, and checks with you before it acts on your behalf," Anthropic says in its announcement. The target is "professionals and small business owners with more on their plates than they can track."

Each morning it sends a short brief on the most important items across your connected apps. It checks those accounts every few hours, not constantly, because nonstop checking would burn through your plan's limits, and it messages you only when something needs you. You can also hand it work, from scheduling meetings to building dashboards, sales forecasts, prototypes and web apps. It brings over the memory, skills and connected apps from your Claude account, and it can tap the more than 6,000 connectors and plugins in the Claude Marketplace. Anthropic's examples include chasing unpaid invoices in QuickBooks or Stripe, preparing payroll in Gusto and sending contracts through Docusign.

The guardrails

Before Duet sends a message, makes a payment, signs a document or deletes anything, it shows you exactly what it's about to do and waits for approval. Some things are off-limits even if you approve them, such as trading securities or moving money between accounts. For purchases, it uses Stripe's Link wallet to pay with a virtual card, so it never sees your real card number. Anthropic says Duet never directly sees your passwords or sign-in tokens either, and that a new classifier blocks it from sharing the most sensitive data, like credit card numbers and government ID. Red teams and an outside firm penetration-tested it.

Anthropic is upfront about the rough edges. Duet "might misread a thread or miscalibrate what's urgent," it's weaker at personal errands like booking tickets, and it can't yet show you its browser. Background checks also count against your usage limits, and big jobs like building a financial model take more.

Why it matters

The fight among AI labs has moved from chatbots to agents that act for you. Meta's Muse launched in September and topped the App Store. Zuckerberg said on October 21 that more than 10 million people in the US and Canada have set it up. OpenAI introduced its "always-on" dots agents weeks later, CNBC notes. Anthropic is aiming Duet at work rather than everyday consumers, in line with its long push for business customers. Team and Enterprise plans don't get it yet, though Anthropic says they will later. The bet is that a cautious agent, one that checks in before every consequential step, is something a small-business owner will actually trust with their inbox and invoices. For now Duet lives apart from the main Claude app. Anthropic says what it learns from early users will shape how the two work together.

Sources: Anthropic: Introducing Claude Duet, Claude Help Center, CNBC.