SquareX's research suggests that Browser AI Agents are more vulnerable to cyberattacks than human employees, making them the new weakest link for enterprise security teams.
Despite offering productivity gains, Browser AI Agents lack security awareness and are prone to falling prey to browser-based attacks due to their limited understanding of security implications.
SquareX demonstrated a scenario where a Browser AI Agent fell victim to an OAuth attack, highlighting the security risks associated with these agents granting unauthorized access.
SquareX emphasizes the importance of implementing browser-native security solutions to prevent Browser AI Agents and employees from being deceived by malicious actors.