A report from SquareX reveals that Browser AI Agents are now a significant cybersecurity vulnerability, surpassing human employees. These agents lack the intuition to detect threats, making them susceptible to attacks, and traditional security measures struggle to identify compromised AI behavior.