Skip to main content
  1. Home
  2. Emerging Tech
  3. News

Anthropic’s Cowork turns Claude into your hands-on digital teammate

Automate mundane tasks without writing a single line of code.

Add as a preferred source on Google
Claude Cowork featured.
Anthropic

Anthropic has announced a new tool that brings Claude Code’s advanced capabilities to less tech-savvy users, letting them perform various actions without writing a single line of code. Dubbed Cowork, the tool can access folders on a user’s computer and read, modify, or delete files on their behalf.

Claude Cowork can also spin up new projects using information from the folder, enabling tasks like generating reports based on a user’s notes. Anthropic says the tool can even be used for organizing old receipts and creating an expense-tracking spreadsheet, or cleaning up a messy downloads folder.

Cowork can also utilize Claude’s existing connectors, allowing it to create documents, presentations, or other projects based on information from third-party apps. In addition, it works with the Claude in Chrome extension to complete actions that require browser access.

Recommended Videos

Anthropic explains Cowork is “designed to make using Claude for new work as simple as possible.” The company adds that users don’t need to repeatedly provide context or manually convert output into the right format, and can queue multiple tasks and let Claude work through them in parallel. The experience is intended to feel more like working with a teammate than having a constant back-and-forth with a chatbot.

The trade-offs of deeper access

The company notes that Cowork’s ability to access local files introduces potential risks. To address privacy concerns, it says that the tool can only read or edit files that users explicitly grant access to. However, since Cowork can delete local files if instructed to do so, Anthropic advises users to be precise with their prompts to reduce the risk of unintended actions.

It also warns about the risk of prompt-injection attacks, which could be particularly problematic if Cowork has access to sensitive files. While Anthropic says it has built safeguards to mitigate such threats, it acknowledges that protections against prompt injection are “still an active area of development.”

Cowork is currently rolling out as a research preview for Claude Max subscribers on macOS. Users on other plans or platforms can join a waitlist to get access in the future.

Pranob Mehrotra
Pranob is a seasoned tech journalist with over eight years of experience covering consumer technology. His work has been…
AI-fueled health policies are depriving the needy in one of the world’s poorest countries
A flawed AI system is driving up medical costs for Kenya’s poorest families
Two kids in Kenya

Kenya was promising wider access to more affordable healthcare. But a new investigation has revealed that its algorithm-driven system is making life harder for the people it was supposed to help most. According to reporting by The Guardian, Africa Uncensored, and Lighthouse Reports, Kenya's new Social Health Authority system is using a predictive machine-learning algorithm to estimate how much people should pay for public health insurance.

The system was first launched back in October 2024 as part of President William Ruto’s promise to expand healthcare access to Kenya's large informal workforce.

Read more
The legendary Oak Ridge lab just developed a portable device that detects GPS-spoofing live
Thieves can hijack your GPS and make everything look perfectly normal. Scientists just built the fix.
GPS spoofing example image

We trust GPS like we trust gravity. It just works and gets us where we want to go. But what if someone could trick it into lying to you, and you'd have absolutely no idea?

Unlike jamming, which floods your GPS with noise and at least lets you know something is wrong, spoofing sends fake signals that look completely legitimate. You might be tracking your car or a shipment and think everything is alright when actually the shipment has been routed to someplace else, with you none being the wiser. 

Read more
I use AI everyday — here are 3 reasons why I paid for Claude over ChatGPT
Claude login screen shown on iPhone

I use AI every single day, so I needed something I could actually depend on, not just occasionally dip into. At some point, it became clear that if I wanted that kind of consistency, I’d have to pay for it. The real confusion started when I had to choose. It came down to ChatGPT and Claude. I’ve used ChatGPT for a long time, and it already understands how I think and what I need, which made it a comfortable choice. But the more I looked into what Claude could do, the harder that decision became. It wasn’t an obvious pick anymore.

I went back and forth for a while, weighing familiarity against capability. In the end, I decided to go with Claude. And in hindsight, I don’t regret that choice one bit.

Read more