Anthropic has introduced Claude Duet, a proactive AI agent designed to monitor connected applications, prioritize tasks, and complete work in the background even when devices are switched off. The tool targets professionals and small business owners and is powered by Claude Opus 5.5. It is available as a research preview through a waitlist for Claude Pro and Max subscribers in the United States.
A Proactive Approach to Daily Work
Unlike conventional chatbots, Claude Duet identifies work that needs attention without waiting for a prompt. It sends users a daily briefing covering important items across connected apps such as email and calendars, and provides updates when tasks are completed or require a decision. Users can assign tasks through text or voice messages, ranging from scheduling meetings and drafting email replies to building dashboards, financial forecasts, and web applications.
The agent connects with more than 6,000 tools and plugins available through the Claude Marketplace, including Gmail, Slack, QuickBooks, Stripe, Gusto, LinkedIn, DocuSign, HubSpot, and Shopify. For instance, it can identify unpaid invoices through QuickBooks or Stripe and draft payment reminders. It can also prepare payroll information, draft job postings, create follow-up emails using customer and deal histories, and help businesses set up and manage online stores.
How Claude Duet Operates
Claude Duet runs on its own cloud-based virtual machine with a browser, allowing it to continue working even when users close the app or switch off their devices. It checks connected accounts every few hours rather than continuously, partly to limit subscription usage. The agent can also search past conversations and use existing Claude memories, skills, and connected apps.
For business purchases, Claude Duet can use Link, Stripe's digital wallet, and Shop Pay to complete transactions with user approval. Link relies on a virtual card so the user's actual card number is not shared, and integration with 1Password for saved website logins is planned. Anthropic also requires approval before the agent sends messages, makes payments, signs documents, or deletes information.
Safety, Approvals, and Early Limitations
Anthropic said Claude Duet includes safeguards designed to prevent unauthorized actions, prompt injection, and sensitive data leaks. These protections include an automated system that assesses proposed actions and controls that keep passwords and sign-in tokens out of the agent's direct view. A separate message-checking system blocks requests to share highly sensitive information, including credit card numbers and government identification, while health and personal finance details can be shared only on explicit user instructions.
Claude Duet has undergone adversarial testing and external penetration testing, and it is covered by Anthropic's public bug bounty program. However, Anthropic acknowledged that the early version may misinterpret conversations, misjudge urgency, or miss information arriving between checks. It is currently better suited to work assignments than personal errands, and users cannot yet view its browser.
Competitive Landscape and Early Access
The launch adds to growing competition in always-on AI agents that work across apps with limited human intervention, including OpenAI's Dots, Meta's Muse, and Bengaluru-based Emergent's Wingman. As agents gain access to more business systems, their permissions and accountability are coming under scrutiny. Recently, CERT-In warned about unauthorized AI agent access to Indian organizations' systems, highlighting the risks associated with deploying increasingly autonomous tools.
Anthropic said feedback from early users will inform further development, and the research preview does not currently include Team and Enterprise subscribers. The company is positioning Claude Duet as a partner that can help professionals and small business owners focus on higher-value work while routine tasks proceed in the background. As the preview expands, its safeguards, judgment, and integration with the broader Claude experience will likely evolve.