Claude Cowork: Anthropic's AI Agent That Actually Works

I've tested a lot of agents over the past couple of years. These experiences expose a consistent pattern of generative AI startups overpromising and underdelivering when it comes to these "agentic" helpers — programs designed to take control of your computer, performing chores and digital errands to free up your time for more important things. But the bots I installed on my laptop would struggle to complete even basic tasks. They just didn't work.

This poor track record makes Anthropic's latest agent, Claude Cowork, a nice surprise. When tested, it worked fairly well — especially for software that's still in beta.

What is Claude Cowork?

Claude Cowork is Anthropic's new general-purpose AI agent that can manipulate, read, and analyze files on a user's computer, as well as create new files. The tool is currently available as a "research preview" only to Max subscribers on $100 or $200 per month plans.

The tool, which the company describes as "Claude Code for the rest of your work," leverages the abilities of Anthropic's popular Claude Code software development assistant but is designed for non-technical users as opposed to programmers.

From Developers to Everyone

Many have pointed out that Claude Code is already more of a general-use agent than a developer-specific tool. It is capable of spinning up apps that perform functions for users across other software. But non-developers have been put off by Claude Code's name and also the fact that Claude Code needs to be used with a coding-specific interface.

"We tried a bunch of different ideas to see what form factor would make sense for a less technical audience that doesn't want to use a terminal," says Boris Cherny, Anthropic's head of Claude Code.

What Can Claude Cowork Do?

Some of the use cases Anthropic showcased for Claude Cowork include:

• Reorganizing downloads — Automatically sort and organize files in your downloads folder
• Expense tracking — Turn receipt screenshots into expense spreadsheets
• Document creation — Produce first drafts from notes across a user's desktop
• Browser automation — Take over the browser to search the web or tidy up a Gmail inbox
• File conversion — Convert file types and generate reports

Anthropic has described the tool, which can work autonomously, as "less like a back-and-forth and more like leaving messages for a coworker."

Built in Just Over a Week

Anthropic reportedly built Cowork in approximately a week and a half, largely using Claude Code itself, according to the head of Claude Code, Boris Cherny. For the past two months, Cherny has written all of his code with AI. Cowork was built using AI tools — a testament to how far AI-assisted development has come.

A Bigger Market Than Claude Code

A top Anthropic executive expects the company's general-purpose AI agent, Cowork, to reach a wider market than Claude Code, the hit product that helped turn the startup into an AI juggernaut.

"This is a general agent that looks well positioned to bring the wildly powerful capabilities of Claude Code to a wider audience," said Simon Willison, a UK-based programmer. "I would be very surprised if Gemini and OpenAI don't follow suit with their own offerings in this category."

Security Considerations

While Cowork works autonomously and can access files on your computer, Anthropic has built in safety measures. The agent operates within defined permissions and users can monitor its activities. However, as with any AI agent that has system access, users should be mindful of what data they allow the tool to access.

The Future of AI Agents

Claude Cowork represents a significant shift in how we think about AI assistants. Instead of just answering questions or generating text, these agents can actually perform tasks on your behalf. This aligns with the broader industry trend toward "agentic" AI — systems that can take action, not just provide information.

For years, AI companies have promised agents that could revolutionize productivity. But most have failed to deliver. Claude Cowork appears to be one of the first that actually works as advertised — at least for file management and basic computing tasks.

The question now is whether the rest of the industry can catch up — and whether users will trust AI agents with more sensitive tasks.

---

This analysis is based on reports from WIRED, Fortune, and Anthropic announcements.