Anthropic launches Claude Duet, a proactive AI assistant for US paid users
Anthropic introduced Claude Duet, an AI assistant that scans emails, calendars, and other connected apps every few hours and alerts users to potential actions—before they ask. The app, available via a waitlist to paid plan users in the U.S. only, summarizes key information daily, drafts replies to missed emails or unpaid invoices, schedules meetings, and forecasts sales. It integrates with over…
Key points
- Claude Duet scans emails, calendars, and 6,000+ external services hourly for users in the U.S. on paid plans
- Requires human approval for actions like sending emails, payments, or deletions to prevent unauthorized steps
- Targeted at small business owners and professionals to automate reminders, invoices, and meeting prep
Safety is a focus: Duet requires human approval before sending emails, making payments, signing documents, or deleting items. It explicitly blocks stock trading, money transfers, and handling card or ID numbers without explicit user consent. However, Anthropic acknowledges the app may misjudge urgency, calling it an experimental feature. Japan remains excluded from the rollout, and no official release date has been set. The company suggests users start identifying tasks they’d delegate to AI now, using prompts to clarify boundaries around automation and oversight.
Model page: Claude Duet →
What is Claude Duet? An AI assistant that acts proactively
note.com · 27 October 2026
Loading the full article…
This text was published by note.com and written by Y2FACTORY AI研究所. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- Anthropic study finds AI agents sway experts’ decisions 86% of the time · 4 src
- Google expands Gemini and Search integrations to third-party apps · 1 src
- Microsoft opens Copilot harness to developers and enterprises via Foundry · 3 src
- Surge AI releases dayjob: Tax benchmark testing AI agents on 50 real tax tasks · 1 src
- Plato raises $16M Series A to build AI agent training simulations · 1 src
Comments
via GitHub Discussions