Browser agents
Also called: computer-use agents · web agents
Browser agents are AI agents that operate a web browser for a user: reading pages, clicking, filling forms and completing tasks across sites. Every page they visit is untrusted input, so they are exposed to indirect prompt injection while often holding logged-in sessions.
How it works
A malicious page, ad, review or email shown in the browser can carry instructions the agent follows, such as visiting another site, entering data or buying something. Because the agent acts with the user's cookies and sessions, a hijack can reach accounts on other sites.
How to defend
Use separate browser profiles without sensitive sessions, require confirmation for purchases, logins and form submissions, and limit the sites an agent may visit.
Real cases 0 stories
No story is filed under this term yet. New ones land here automatically.