# Security
<a name="security"></a>

The security portion of this lens provides security best practices and questions for organizations implementing autonomous AI agents and agentic systems on AWS. This focuses on securing AI agents, protecting against prompt injection, maintaining data privacy, and maintaining control over autonomous operations.

**Capabilities**
+ [Secure agent memory and state](agentsec01.html)
+ [Secure agent tool usage](agentsec02.html)
+ [Agent identity and permission management](agentsec03.html)
+ [Agent goal alignment and manipulation prevention](agentsec04.html)
+ [Agent observability and non-repudiation](agentsec05.html)
+ [Secure multi-agent orchestration](agentsec06.html)
+ [Human oversight protection and agent containment](agentsec07.html)
+ [Secure agent inputs and outputs](agentsec08.html)
+ [Agent vulnerability scanning and penetration testing](agentsec09.html)