Malicious browser extension could drive Claude Desktop to run code via Dispatch
Anthropic has reportedly said Dispatch will be retired, yet its help page still tells existing users they can keep using it.
Security firm Forever Security showed that a malicious or compromised browser extension running on claude.ai could drive Claude Desktop into running commands on a user's machine, using the Dispatch feature that lets people send tasks from a phone. In their test, the extension sent Claude prompts that led it to install an npm package, and the install script opened the Calculator app; the researchers say the same route could have stolen files or planted a backdoor. Dispatch accepted an ordinary signed-in claude.ai session without showing any link to it, and the researchers reported the flaw to Anthropic on May 14, which accepted it but closed the report as a duplicate. Anthropic reportedly told the researchers on October 13 that Dispatch would be retired by the end of October, though its help page still says existing users can keep using it. Developers and companies should disable cross-device Dispatch if unneeded, scrutinise browser extensions, and monitor AI agents for unexpected installs or credential access.
Why it matters
The test shows that anyone with a compromised browser extension on claude.ai could have an AI agent on another device act on their behalf without a visible link or any approval, which matters to ordinary users who rely on the feature. The report was accepted but closed as a duplicate, and the planned retirement leaves unclear whether the weakness will be fixed or simply left in place as users are moved away. Anyone who uses Dispatch is left depending on that timeline rather than on a confirmed repair.
Matters or hype?
Does this story matter, or is it hype? Vote before you see what everyone else said.