Anthropic launches Claude Duet AI agent for professionals
Anthropic has entered the competitive AI agent market with Claude Duet, a proactive tool for professionals and small businesses to automate tasks and manage workflows, launching as a research preview in the US.

Anthropic has launched Claude Duet, a new AI agent designed for professionals and small business owners. The move signals the industry's broader shift from passive chatbots to autonomous agents that can act on a user's behalf across connected applications. Claude Duet enters a crowded field dominated by rivals like OpenAI's Dots, Meta's Muse, and Google's Workspace ambitions.
The market is exploding because users have hit a ceiling with current chatbots, tired of manually copying and pasting between AI interfaces and their work tools. Agents promise to break this barrier by living inside workflows, understanding context across apps, and executing multi-step tasks autonomously. However, this new phase faces friction, with companies like Yelp and Delta reportedly blocking autonomous agents from some providers over data scraping and terms of service concerns. Anthropic arrives late to the party but carries a reputation for safety and a model architecture trusted by many developers.
Claude Duet's core capabilities
Claude Duet is an early research preview, available through a waitlist for users on the Claude Pro and Max paid plans in the United States. It is accessible as a standalone app on iOS and Android, plus through the existing Claude web interface. Anthropic designed the agent around its flagship Claude Opus 5.5 model, balancing capability with careful reasoning.
The agent functions as a proactive productivity partner. Each morning, it sends users a curated brief highlighting important items across connected apps. Throughout the day, it monitors those apps and only surfaces notifications when something requires human attention. Users can hand off entire tasks by text or voice, from scheduling meetings to building dashboards and sales forecasts. Claude Duet completes tasks in the background using its own virtual machine, allowing a user's device to be off, and will return for approval if needed.
Anthropic distinguishes its offering from personal agents. "Duet isn’t good at handling errands or booking tickets, and it can’t show you its browser use as it happens," the company said. Instead, it is positioned as a genuine work partner. The agent can connect to over 6,000 connectors and plug-ins in the Claude Marketplace, including Gmail, Slack, QuickBooks, and Shopify.
Safety and commercial actions
A key feature is the agent's ability to execute commercial actions. "It can also make online purchases on users' behalf via Stripe's Link digital wallet or Shopify's Shop Pay," Anthropic stated. The company emphasized that sensitive steps like payments, sending messages, signing documents, or deleting data would always require explicit user approval first. Claude Duet will not trade securities or move money between accounts.
On safety, Anthropic released a new classifier to check each message to prevent sensitive information from leaking out. By default, the agent would turn down requests for credit card numbers and government identification. Anthropic says Claude Duet never directly sees user passwords, and sign-ins using logins saved in 1Password are coming soon.
The competitive landscape and rollout
The launch comes as many of Anthropic's rivals focus on business use cases. While Meta's Muse is a personal assistant, the company also released a version for small businesses. OpenAI has been testing Dots with select users for months, with reports suggesting deep integration with Microsoft's ecosystem. Google has its own agent ambitions tied to Workspace.
Anthropic has positioned itself as the responsible AI player, recently committing 5% of equity to a public benefit fund. Its cautious rollout of Claude Duet as a research preview buys time to refine the experience amid the competitive race. The company has not disclosed whether the agent will be bundled with existing subscriptions or priced separately, but premium pricing is expected given the compute costs of continuous monitoring.
This cautious approach reflects broader industry recognition that businesses need guardrails before handing AI the keys to their data. Nvidia recently launched OpenShell to help companies deploy agents safely on edge systems. For professionals drowning in app overload, Claude Duet represents a promise of AI-powered relief. Its performance in this research preview across iOS, Android, and web platforms will test whether Anthropic's reputation for safety and reliability can translate into market share.





