The time to hack away at a laptop to keep AI agents running is coming to an end. On Tuesday, Anthropic announced that Claude Cowork, its AI agent designed to handle digital tasks for you, is expanding beyond the desktop app.
You don’t have to leave your device running, laptop open, to keep the agent running and running scheduled tasks overnight. Anthropic also announced limited versions of Cowork that users can interact with through its existing Claude smartphone app or web browser, without the desktop connection previously required.
In the feature’s launch video, Anthropic shows someone asking for help with a business deal renewal scheduled for the next day. In one go, a user asks Cowork to aggregate data from email threads, Slack channels, meeting transcripts, and recent online chat. The user then asks Cowork to use that information to generate a meeting reference document and a pre-written email. Cowork previously could do all that while your desktop session was active, at least. Now, the agent can run even after turning off the clock, catching those late-night incoming messages.
I first tried Claude Cowork when it dropped in January, and I was immediately amazed that this agent could follow and complete the tasks I asked it to run on my laptop, such as organizing a mess of screenshot files into decent, labeled folders. It also does a solid job of helping me schedule events on my calendar. The agent wasn’t perfect, and it still exposed me to the risk of instant injection or other security breaches, yet Cowork felt like a game changer in how everyday users could use their devices.
This is not the first time Claude users have been able to interact with Anthropic agents on their mobile devices. Previously, users could pair their smartphone app with their desktop using the Dispatch feature. This allows users to send job requests from their phone, wherever they are. But this method had one major limitation. “Your computer must be awake, and the operating system must be open for Claude to run the tasks,” reads Anthropic’s description. That’s why some users leave their laptops on to keep sessions running. Now, Cowork can run tasks without an active desktop session.
The announcement is part of a big, recent shift in Silicon Valley toward automated, autonomous AI agents that you can control by sending a text. This trend was inspired by OpenClaw, a homebrew agent with a lobster mascot that went viral in early 2026, as the first adopters ran it 24/7 and gave it control over their online lives.
Some tech companies were jealous of all this crustacean-centric praise. Therefore, in the first half of the year, OpenAI hired the creator of OpenClaw and launched Codex, its flexible agent; Google introduced Spark, its take on an always-on agent; and Anthropic became more dependent on making its agents more user-friendly. Anthropic’s biggest hit was Claude Code, which helps developers automate tasks. Cowork takes the same approach, translates it out of the context of a computer terminal, and puts that power into chatbot form for regular users.
Anthropic plans to release this enhanced version of Cowork as a beta to subscribers of its Max plan, which starts at $100 per month. After that, the features are expected to trickle down to members of Anthropic’s cheapest tier, Pro, which costs $20 a month. It is unclear if this will be released to free users, who do not have access to Claude Cowork in their subscriber section.