---
title: "The safety of AI agents is determined outside the model"
date: 2026-09-19
source: http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6aaf5810920f477bb3651adde0e49dba&url=https%3a%2f%2fnote.com%2fquietship%2fn%2fn96a2a7b7522c%3fhl%3den&c=15406634408073783769&mkt=en-us
description: "IntroductionWhen operating AI agents, the question arises of how to use them safely. A common approach is the idea of properly instructing the model. You write in the instructions: 'Do not do anything ..."
---

# The safety of AI agents is determined outside the model

IntroductionWhen operating AI agents, the question arises of how to use them safely. A common approach is the idea of properly instructing the model. You write in the instructions: 'Do not do anything ...

*Published: 2026-09-19*