Anthropic released a personal AI agent called Claude Duet on October 21, making it the latest AI lab to offer an always-on assistant after SpaceXSI, Meta and OpenAI. The company says the agent can carry out tasks on a user's behalf even when that person is away from their phone or computer. Duet operates on its own virtual machine and has access to a dedicated browser it can use to surf the web and finish jobs. It can also draw on any apps connected to a user's Claude account.
Anthropic says Duet will be included as a perk of its $20 per month Pro plan and higher tiers, though the company notes that could change. That pricing sets it apart from OpenAI's Dot agents, which are currently available only through OpenAI's own Pro plans. Those plans start at $100, and OpenAI recently added a $500 per month option. Anthropic is not yet offering Claude Duet to Team and Enterprise customers.
During the preview period, Anthropic says Duet will share usage with a customer's other Claude apps, including Code and Cowork. The company said it built the agent to check connected accounts every few hours rather than continuously, because checking nonstop would exhaust a plan's limits much faster. Anthropic described the preview as a way to gather feedback on how often Duet messages users before a wider rollout.
Anthropic is marketing Duet as a proactive agent that can produce a daily briefing with a list of action items drawn from the apps linked to a Claude account. Users can also hand it tasks such as scheduling meetings, creating forecasts and building web apps. As one example, a business owner could ask Duet to generate a weekly sales and cashflow dashboard, while a scientist could have the agent flag newly published research papers in their field.
According to Anthropic, the difference from the standard Claude app is that Duet initiates contact rather than waiting for a chat to be opened. The company says the agent reports what needs attention each morning, accepts tasks sent by message, and reaches out when a job is finished or requires approval.
Duet's proactive behavior depends on data in a user's Claude account, including information the chatbot has stored in its memory and any apps connected to the Anthropic account. How useful the agent proves will therefore hinge on how much of their digital lives users are willing to share with Anthropic. For some customers, that privacy tradeoff may not be worth it.
More AI news from TechManNews.
