Researchers from an Israeli cybersecurity firm have identified a significant vulnerability affecting popular AI browser assistants. The flaw allows attackers to manipulate these assistants into displaying fake customer service numbers, fraudulent links, and even extracting sensitive user data without consent. Users may unknowingly have their login credentials stolen or be exposed to false information that could influence their decisions. The affected AI assistants include those from major technology companies, raising concerns about the security of widely used digital tools.