Malicious browser extensions exploit Claude Dispatch for remote code execution
Original title: Malicious Browser Extensions Could Use Claude Dispatch to Run Code on Users' Computers
Researchers demonstrate extensions sending prompts via Claude Dispatch to run npm installs on connected computers; Anthropic plans to retire the feature.
Researchers at Forever Security showed malicious browser extensions can exploit Claude Dispatch to send prompts to Claude Desktop, causing it to install packages and run commands like opening the Calculator app on users' computers. The attack works by accepting a signed-in session from claude.ai and automating approvals in the browser. Anthropic confirmed the flaw but plans to retire Dispatch by the end of October while similar features remain a risk.
Unconfirmed: Anthropic closed the report as a duplicate of an internal issue
- Dispatch accepts unsigned sessions from claude.ai
- Claude Desktop runs global npm install and Calculator app
- Attack works across devices signed to same account
- Anthropic retired feature by end of October
- Researchers reported flaw in May, closed as duplicate
Impact: Remote code execution via browser extension on Claude Desktop
- 1.Turn off cross-device Dispatch
- 2.Review device pairings and approvals
- 3.Monitor AI agent actions on computer
- Vendors
- Anthropic
- Products
- Claude Desktop
- Organizations
- Forever Security
- AI models
- Claude
From the original article
3 min read at thehackernews.com
A browser extension that can run scripts on claude. ai could make Claude Desktop run code on a person's computer, researchers at Forever Security, an endpoint security company, showed in a test. The attack needs Dispatch, a Claude feature that lets people send tasks from their phone to Claude Desktop.
thehackernews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.