Re:Optimizing Content for Agents
by tanbablack on Saturday March 14, 2026 @11:46AM
(#47377848)
Great writeup. Attackers are also "optimizing content for agents" — just with malicious intent.
Unit42 published research in March 2026 confirming websites in the wild embedding hidden instructions specifically targeting AI agents.
Techniques include zero-font CSS text, invisible divs, and JS dynamic injection. One site had 24 layered injection attempts.
The same properties that make content agent-friendly (structured, parseable, in the DOM) also make it a perfect delivery mechanism for indirect prompt injection.
Unit42 published research in March 2026 confirming websites in the wild embedding hidden instructions specifically targeting AI agents. Techniques include zero-font CSS text, invisible divs, and JS dynamic injection. One site had 24 layered injection attempts.
The same properties that make content agent-friendly (structured, parseable, in the DOM) also make it a perfect delivery mechanism for indirect prompt injection.