Prompt injection
In shortHidden instructions inside a web page, document or tool result that trick an AI agent into doing something its user never asked for.
Where it sits in EndLayer · step 6 of 7, decision
What it means
An AI agent that reads untrusted content can be steered by it. Limiting autonomy and reach, and stopping bad actions with a guardrail, reduces the damage.
Why it matters
It is the most common way AI agents are turned against the people using them.