Researchers demonstrated that a malicious or hijacked browser extension can use a logged-in Claude web session to send instructions to Claude Desktop and run commands on a computer. The proof of concept repeatedly pressured the AI agent to install a malicious npm package, while automated clicks approved prompts the user never saw. Shared Claude accounts can extend the risk from a personal browser to a corporate laptop.
Mirage safely runs attacks like this one against your own team, so you find out what happens before a real adversary does.
Your personal Claude tab could send commands to your work laptop. A compromised browser extension can use your logged-in Claude Dispatch session. It says you're AFK, asks to list folders, then pushes Claude Code to install add-colors-to-bash. Here's the catch: Dispatch connects devices on one account. Researchers kept prompting and automatically clicked Allow, so the work laptop could run commands while its owner wasn't watching. One move: keep your personal Claude account disconnected from your work laptop.
Researchers blocked an attempted attack in which a malicious instruction file in a Downloads folder tried to make OpenAI Codex install malware. Tests showed that similar files could manipulate coding agents into weakening safeguards, changing system-wide settings, redirecting API traffic, and…
Researchers found that a malicious instruction file placed in a user’s Downloads folder can manipulate AI coding agents into installing malicious software or exposing account login tokens. In testing, OpenAI Codex followed the planted instructions more than 90% of the time, while a related attack…
Researchers stopped a real attack in which a malicious instruction file in a developer’s Downloads folder attempted to direct an AI coding agent to install known malware. The attack abuses the agent’s trust in local instruction files and can also alter agent settings to steal session tokens or make…
North Korean operatives reportedly posed as remote IT workers, used stolen identities, passed hiring checks, and gained legitimate access inside hundreds of Western companies. The article also describes attackers using autonomous AI agents inside a victim’s cloud environment to harvest credentials…
Researchers demonstrated that malicious web pages and local previews can abuse embedded browsers in popular AI coding assistants. The attacks can hide instructions from users, spoof trusted login addresses, or access sensitive files, creating realistic risks for developers who open untrusted links…
Researchers found a malicious instruction file in a customer’s Downloads folder that attempted to manipulate an AI coding agent into installing a rogue npm package. The technique can work because browsers may save files into Downloads without prompting, and coding agents may treat nearby…
