Security News

Researcher Says Claude Dispatch Flaw Lets Browser Extensions Run Commands on Your Computer, and Anthropic Hasn't Fixed It

The Register · 5 Nov 2026

Key Takeaway Review and limit the browser extensions installed on staff devices, and be cautious about enabling AI agent features that can run commands on a computer, especially where those devices are signed into shared or business accounts.

Forever Security researcher Gal Weizman has found a vulnerability, dubbed ThisBash, in Anthropic's Claude products. According to The Register, any browser extension that can control Claude Dispatch can make Claude Desktop run code on the computer, or on a different machine signed into the same Claude account. No browser exploit is needed and user interaction is minimal.

Dispatch lets users send and manage tasks from a phone or browser to Claude Desktop, an AI agent that runs commands on their computer. Weizman found that Dispatch's interface and API could be reached from an ordinary logged-in web session at claude.ai. An extension able to control that session could therefore send commands to Claude Desktop as if they came from the legitimate user. Weizman said this gives attackers a way to compromise a machine completely.

Anthropic acknowledged the report as valid in May but closed it as a duplicate of a known issue. It paid no bug bounty, issued no CVE, and told Weizman it planned to retire Dispatch at the end of October. As of October 30, its support page still listed Dispatch as available in limited beta for some plans. Weizman says the underlying design weakness remains and now affects other features, including Claude Code Remote Control and Claude Cowork sessions. Anthropic did not respond to The Register's questions.

Anthropic Claude AI agents browser extensions vulnerability

Summarised by CISO AI from The Register, written with Claude Sonnet 5.5. We link back to every original so you can read it yourself.