Anthropic has launched Claude Duet, a new artificial intelligence agent designed to manage tasks proactively, as the company competes in the growing market for AI assistants that work on users’ behalf.
Claude Duet is aimed at professionals and small business owners who need help organizing work, managing connected applications and completing tasks with minimal instructions. Unlike traditional chatbots that primarily respond to prompts, Claude Duet can monitor connected apps, identify issues that need attention and continue working in the background.
READ: Anthropic CEO Dario Amodei calls India’s AI Summit ‘disorganized’ (June 19, 2026)
The agent is powered by Anthropic’s Claude Opus 5.5 model. Each morning, it prepares a briefing highlighting important developments across connected accounts. It continues checking those applications throughout the day and notifies users when a decision or intervention is needed.
According to Anthropic, the agent is designed to prevent sensitive information from being shared without authorization. Details about health or personal finances are shared only when users explicitly instruct Claude Duet to send them. The agent can also help users make online purchases through Link, Stripe’s digital wallet, and Shopify’s Shop Pay.
“Claude Duet confirms the item with you, and Link asks you to approve the spend. It then pays with a virtual card, so your real card number isn’t shared,” Anthropic said in its blog post.
Users can assign the agent a range of tasks, from scheduling meetings and drafting follow-up emails to building dashboards, preparing financial forecasts and developing web applications. Claude Duet operates through its own virtual machine and browser, allowing it to continue working even when users step away from their devices.
The agent can also connect to more than 6,000 connectors and plugins in the Claude Marketplace, enabling it to work with a range of business applications.
Claude Duet is available as a research preview through a waitlist for subscribers to Anthropic’s Pro and Max plans in the U.S. The standalone app is accessible on iOS, Android and the web through Claude.ai.
Anthropic said the product includes safeguards designed to keep users in control of important decisions. Claude Duet seeks approval before sending messages, making payments, signing documents or deleting information. By default, it rejects requests to share highly sensitive details, including credit card numbers and government identification. Some actions, such as trading securities and transferring money between accounts, remain prohibited even with user approval.
READ: OpenAI, Anthropic and SpaceXAI leaders back calls to slow AI development (September 16, 2026)
The launch comes as AI companies increasingly compete to develop agents that can handle multiple tasks and reduce users’ workloads, rather than simply generate text or answer questions. Claude Duet enters a growing market for proactive AI assistants, alongside products from other major technology companies.
However, as AI companies develop more sophisticated agents, concerns remain about data breaches, unauthorized actions and security vulnerabilities. One reported incident involving Meta’s Muse AI raised questions about the risks of autonomous agents after the system reportedly negotiated a Facebook Marketplace deal, shared a user’s address and arranged a late-night pickup without the user’s approval.