AutoJack exploit chain allows web-enabled AI agents to trigger host-level RCE (opens in new tab)

Microsoft researchers have identified a critical exploit chain named AutoJack that allows malicious web content to compromise the host machine running an AI agent. The vulnerability specifically targeted AutoGen Studio, a prototyping interface for multi-agent systems, by abusing its trusted local access. By chaining three separate weaknesses, an attacker can bypass localhost security boundaries to execute arbitrary commands under the developer's account. <a href="