Anthropic is reportedly working to make Claude more of a digital coworker than a chatbot with Claude Duet. The tool is intended to follow what is happening across the work apps a user connects, figure out which jobs require action and help get them done. Its potential uses include coordinating calendars, putting together data dashboards and working on financial projections, with a focus on easing the daily workload of individual professionals and small businesses.
AI Companies Want Their Assistants to Do the Work, Not Just Talk About It #
The race to build more capable AI assistants is moving beyond chat windows and one-off prompts. Tech companies are now exploring tools that can keep track of ongoing activity, pull information from different services and take action without users having to spell out every step. Anthropic is entering this space with Claude Duet, joining Meta and OpenAI in a push to make AI more involved in everyday tasks. Its approach, however, is centred on getting work done rather than managing a user's personal life.
Claude Duet draws on years of our safety and security work on agents. It never directly sees your passwords, a new classifier guards against sharing sensitive information, and the important decisions stay with you. We’ll share more on our approach to safety in Claude Duet soon.
— Claude (@claudeai)October 21, 2026
Claude Duet Aims to Become an AI Coworker #
Claude Duet is designed to help users stay on top of their work without constantly checking multiple apps. It can prepare daily updates, surface information from connected services and assist with jobs such as arranging meetings, compiling dashboards and producing financial projections. The idea is to reduce the time people spend collecting information and coordinating routine tasks. Unlike a conventional chatbot that waits for a question, an AI agent is intended to work through a series of steps to achieve a specific outcome.
Anthropic’s AI Agent Still Needs Human Oversight #
The proposed workflow goes beyond responding to commands: Claude Duet can continue working while a user is away, flag information that may need attention and request permission before carrying out certain actions. However, Anthropic has cautioned that the preview is still being developed. The agent may misread a discussion or fail to recognise which issue deserves immediate attention, making human checks important. It is also geared towards professional productivity rather than every kind of personal errand, so users should not expect it to handle all aspects of their lives independently.
Claude Duet Faces Competition From Meta Muse and OpenAI Dots #
Anthropic is entering a market where rivals are already exploring different versions of always-available AI assistance. Meta's Muse is positioned around personal assistance, while OpenAI's Dots is intended to support ongoing tasks across consumer and work settings. Meta CEO Mark Zuckerberg said during the company's third-quarter earnings call that more than 10 million people in the US and Canada had set up Muse agents in the previous six weeks. That figure indicates early adoption, though it does not reveal how regularly those agents are used. Claude Duet, meanwhile, is available through a waitlist, with preview access limited to eligible Pro and Max subscribers via a standalone app or the Claude website.
ALSO SEE: