It's a busy week for consumer AI agents, and Anthropic just joined the race. The company announced Claude Duet, a personal AI agent that you message like a teammate. Powered by Opus 5.5, Claude Duet handles your inbox, calendar and follow-ups, works on its own cloud computer, and messages you when something needs your attention or your approval.
The launch arrives right after the Meta Muse and OpenAI Dots moments that already put proactive AI agents on everyone's radar. Claude Duet is currently a research preview on a waitlist, available in Pro and Max plans in the US, so Tunisian users can't access it yet. For Anthropic, it's a clear signal that the next AI fight is not about chat, but about who owns your daily workflow.
What Claude Duet actually does
Unlike the standard Claude app, which waits for you to open a chat, Claude Duet comes to you. It can send a morning brief that flags what needs you in your inbox and calendar, then draft replies for you to approve. It also takes on larger projects: finding unpaid invoices, drafting reminders, preparing payroll for review, getting contracts ready to send, or building a brief, deck or dashboard.
Anthropic says it keeps working while you're away and goes quiet until something needs you, waiting for your OK on sensitive operations like sending messages or paying. Approvals, logins and safety checks run in your own cloud computer so your details stay private, and you can see what it has done and is doing so you can steer it.
Claude Duet features and integration
You message Claude Duet in the app through text or voice, with no prompts or setup. Your memory, skills and connectors from the Claude app are already there from the first message, and everything lives in one thread. Finished work comes back as a doc, sheet or deck for your review.
- MESSAGE
