Brijesh Desai 
Anthropic Launches Claude Duet AI Agent. It is a proactive AI agent that monitors apps, manages tasks, and works in the background—challenging Meta’s Muse and OpenAI’s Dots.
Anthropic has entered the personal AI agent race with Claude Duet, a standalone app designed to do more than answer prompts. Launched in research preview on October 21, Duet is built to monitor connected apps, surface what needs attention, complete tasks, and keep working even when the user steps away.
The launch puts Anthropic directly against Meta’s Muse and OpenAI’s Dots—two recently introduced “always-on” AI agents that are trying to turn chatbots into active workplace collaborators. Anthropic is positioning Duet for professionals and small-business owners who need help managing a scattered digital workday.
What Claude Duet Does
Claude Duet starts each day with a short briefing based on activity across a user’s connected apps. It then checks those apps periodically and alerts the user when something requires action.
Users can also assign it work directly. Anthropic says Duet can handle tasks such as scheduling meetings, building financial forecasts, and creating web apps. It can continue working in the background and checks with the user before taking significant actions on their behalf.
The agent runs on Anthropic’s Opus 5.5 model and uses a virtual machine with its own browser to complete tasks. That setup allows it to interact with web-based tools without needing access to the user’s local computer.
How It Compares With Muse and Dots
Meta launched Muse last month as a personal AI agent, and the app quickly climbed Apple’s App Store rankings. OpenAI followed with Dots, its always-on agents designed to pursue goals across apps such as Slack and Microsoft Teams.
Claude Duet’s differentiation is its focus on practical work management. Rather than acting mainly as a conversational assistant, it behaves more like a proactive chief of staff: reviewing connected apps, flagging priorities, and carrying out follow-up tasks.
Duet also has a broad integration base. Anthropic says it connects with more than 6,000 connectors in the Claude Marketplace, including Gmail and Slack. It can even make purchases through Stripe’s Link wallet and Shopify’s Shop Pay—though that capability will understandably raise questions about guardrails and user control.
Availability and Pricing
For now, Claude Duet is available through a waitlisted research preview in the United States. It is initially limited to Claude Pro and Max subscribers and shares their existing usage limits.
The app is available on iOS, Android, and through claude.ai. Anthropic has said Pro subscribers, who pay $20 per month, will get access when Duet reaches general availability.
That pricing approach could give Anthropic an advantage. OpenAI’s Dots are currently limited to its Pro-tier subscribers, while Anthropic plans to bring Duet to its $20 Pro plan at launch.
Why the AI Agent Race Matters
The shift from chatbots to agents is becoming the next major battleground in AI. Instead of waiting for a user to type a request, these systems are designed to observe context, suggest next steps, and execute multi-step work.
That could be genuinely useful for busy professionals. Imagine Duet noticing an unanswered client email, checking your calendar, drafting a response, and scheduling a follow-up meeting—then asking for approval before sending anything.
But the same autonomy creates real concerns. Agents with access to email, calendars, messaging apps, and payment tools need strong permissions, transparent activity logs, and clear limits. Anthropic says Duet checks with users before acting on their behalf, which is a sensible starting point, but trust will depend on how reliably those safeguards work in practice.
Anthropic’s Late—but Careful—Entry
Anthropic isn’t first to market. Meta’s Muse and OpenAI’s Dots arrived earlier, and Muse reportedly attracted millions of downloads.
Still, being late isn’t necessarily a disadvantage. Anthropic can learn from early user behavior, refine its safety model, and position Duet around reliability rather than hype. The company’s emphasis on asking permission before acting may appeal to business users who don’t want an AI agent making irreversible decisions.
Claude Duet is still an early preview, so its real-world usefulness remains to be seen. But the direction is clear: AI companies are no longer competing only to build smarter models. They’re competing to build assistants that can actually run parts of your workday.