Quick Overview
- Anthropic unveiled a capability enabling Claude to autonomously operate Mac computers from a distance
- Subscribers can delegate tasks via mobile devices while Claude handles applications, web navigation, and document management
- The functionality integrates with Dispatch, allowing task assignment to Claude through smartphone interfaces
- Anthropic cautions that the technology remains experimental and recommends keeping it away from sensitive information
- This release follows the viral success of OpenClaw, an agentic AI competitor that sparked industry-wide development efforts
Anthropic introduced a groundbreaking capability on Monday that enables its Claude artificial intelligence to autonomously operate users’ computers for task completion. The functionality is currently accessible to Claude Pro and Claude Max paying members.
[[TWITTER_EMBED]]
Currently limited to MacOS systems, the feature allows subscribers to issue commands from mobile devices while Claude executes the instructions on desktop machines — launching applications, conducting web searches, populating data fields in spreadsheets, or reorganizing files.
During a demonstration, Anthropic displayed how a user delayed for a meeting requested Claude to convert a presentation into PDF format and add it to a calendar appointment. The AI completed the entire sequence independently.
The technology interfaces with services such as Google Calendar and Slack through native integrations. When direct connections aren’t available, Claude resorts to simulating keyboard and mouse interactions to accomplish objectives.
Before accessing any new application, Claude requests explicit authorization. Subscribers maintain the ability to terminate operations at any point during execution.
Integration With Dispatch System
This remote control capability complements Dispatch, a companion tool Anthropic introduced recently within Claude Cowork. Dispatch enables sustained conversations with Claude across multiple devices and facilitates ongoing task delegation.
When combined, these features can manage responsibilities such as delivering morning summaries, executing code validation routines, or monitoring email communications without manual intervention.
The company acknowledges that sophisticated operations may require multiple attempts to complete successfully. Anthropic characterizes this release as an experimental preview designed to collect user feedback and pinpoint areas requiring enhancement.
Privacy Concerns and Safety Measures
[[LINK_START_1]]Anthropic[[LINK_END_1]] has implemented protective measures to minimize vulnerabilities, including automated detection systems for prompt injection exploits. Nevertheless, the organization advises against deploying the feature with applications containing confidential information — certain sensitive apps are restricted by default.
Cybersecurity professionals have raised broader questions about autonomous AI agents. These systems can execute substantial actions rapidly with minimal oversight. Additional concerns include the possibility of malicious actors commandeering the agent to infiltrate private documents or networks.
Anthropod committed to continuously enhancing its security protocols as emerging threats develop.
Competitive Landscape
This launch intensifies Anthropic’s competition with OpenClaw, an open-source platform that gained widespread attention recently. OpenClaw permits users to submit instructions through WhatsApp or Telegram for local device execution. Nvidia recently unveiled NemoClaw, a commercial adaptation of this technology.
OpenAI entered this market segment last month by recruiting Peter Steinberger, OpenClaw’s developer, to contribute to personal AI assistant development.
Currently, Claude’s remote computer control remains available exclusively on MacOS for Pro and Max tier subscribers during its research preview phase.





