Anthropic has extended its Claude Cowork agent from a desktop-only application to web and mobile platforms, rolling out beta access to Max subscribers. This evolution aims to provide uninterrupted workflow continuity whether at a desk or on the go by routing tasks intelligently between coding and knowledge work engines.
- Claude Cowork extends to web and mobile for task continuity.
- Dispatch feature routes workflows to appropriate AI engines.
- Business process tasks dominate over coding in usage data.
What happened
Anthropic launched Claude Cowork originally as a desktop app in January, aimed at facilitating general knowledge work with AI assistance. On July 7, it initiated a beta rollout of the tool's new web and mobile implementations, initially available to Max subscribers. This multi-device expansion allows users to begin tasks on their desktop and seamlessly continue or monitor progress via phone or web browser, even when not at their desks.
The company introduced a new feature called Dispatch that keeps a single persistent conversation thread and intelligently directs different requests to the optimal Claude engine. Development or coding-related requests are processed by Claude Code, while administrative or knowledge-based tasks remain with Claude Cowork. Rather than flooding users with step-by-step details, the system only notifies users when their input or decisions are required.
Why it matters
This move reflects a broader industry trend where effective AI adoption depends on integrating assistant capabilities directly into the environments where work gets done, rather than relying solely on chatbot interfaces. By enabling continuity across multiple devices, Anthropic positions Claude Cowork as an unobtrusive AI coworker that works alongside human users to handle routine tasks and reduce friction in workflows.
Anthropic’s usage data from over 1.2 million sessions reveals that the largest category of AI use is business process tasks (33.4%), followed by content creation (16.4%), with coding representing under 9%. This challenges the narrative of AI assistants primarily serving developers and highlights their increasing adoption for everyday office automation in finance, HR, and administration.
What to watch next
As Claude Cowork expands beyond the desktop, monitoring how users engage with AI agents across devices will be critical, particularly concerning security risks linked to remote control of desktops via phones or web interfaces. Anthropic advises caution by recommending users connect only trusted applications in the workflow to avoid accidental or malicious commands that could cause harm.
In the competitive landscape, Anthropic’s push parallels moves by giants like OpenAI and Google, who are similarly embedding AI assistants in enterprise workflows and popular office software. The success of Claude Cowork’s multi-device strategy could influence broader adoption trends and shape the future design of AI assistants focused on real-world productivity beyond chatbot conversations.