Our new study, "The Hidden Dangers of Browsing AI Agents" offers the first end-to-end threat model for browsing agents and ...