Anthropic Launches Claude CoWorker, an AI Desktop Assistant

Published on January 14, 2026 | Translated from Spanish
Screenshot showing the Claude CoWorker interface overlaid on a text document window, with highlighted editing suggestions.

Anthropic Launches Claude CoWorker, an AI Desktop Assistant

The artificial intelligence company Anthropic has unveiled a new operating mode for its Claude model, called Claude CoWorker. This feature is designed to integrate seamlessly into the daily workflow, acting as a persistent digital collaborator that resides directly on the user's desktop. 🖥️

How Does This Desktop Assistant Work?

The CoWorker mode is activated via a configurable keyboard shortcut. Once started, it operates continuously in the background, with the ability to capture and analyze the content visible on the screen. This allows it to understand the user's immediate context, whether editing a document, reviewing a spreadsheet, writing an email, or browsing the web. Its main goal is to reduce the need to constantly switch between windows and applications.

Main Features of the System:
The ideal assistant that never asks for a coffee break, although sometimes its interpretation of "review this report" can be as literal as suggesting to change the font.

Privacy and Operational Limits

Anthropic emphasizes that processing the screenshot is performed locally on the user's device. By default, it does not send sensitive visual data to its remote servers. The model has been trained to try to skip private information, although the ultimate responsibility for what is displayed on the screen lies with the user. 🔒

Key Aspects of Its Capabilities and Restrictions:

A Step Toward Total AI Integration

With Claude CoWorker, Anthrop

Related Links