2.2 Determine Recommended Standard Architectures for AI Security NETWORK-SECURITY-ARCHITECT Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   A financial institution is deploying an AI system for real-time trading. The system must have extremely low latency, and the security team needs to detect and block prompt injection attacks. The institution also must comply with regulations that require all data to remain within the country. Which architecture should the architect recommend?

   • AInline AIRS with threat intelligence feeds, deployed in a public cloud region in another country
   • BOut-of-band AIRS with behavioral models, deployed in a public cloud region in the same country
   • CCloud-native AIRS with behavioral models, deployed in a local data center
   • DInline AIRS with AI-specific signatures, deployed in a local data center
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   How do AI security content and data security controls complement each other in an AI architecture?

   • AAI security content detects and blocks AI-specific threats, while data security controls protect the confidentiality and integrity of the data itself.
   • BThey are independent and do not need to be integrated.
   • CData security controls are only needed for data at rest, while AI security content handles all other threats.
   • DAI security content replaces the need for data security controls.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   Which type of AI security content is specifically designed to detect known malicious patterns in prompts, responses, or model behavior?

   • AThreat intelligence
   • BAI-specific signatures
   • CBehavioral models
   • DData loss prevention rules
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which type of AI security content is most useful for detecting novel or zero-day attacks that do not match known signatures?

   • AStatic analysis rules
   • BThreat intelligence feeds
   • CAI-specific signatures
   • DBehavioral models
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   A media company is using a cloud-based AI service to generate personalized content. The security team wants to ensure that the AI service is not abused to generate harmful content and that all interactions are logged for compliance. The company has no on-premises infrastructure. Which AIRS form factor should be recommended?

   • AInline, deployed as a virtual appliance in a separate cloud VPC
   • BOut-of-band, using cloud storage logs
   • CCloud-native, integrated with the AI service's API
   • DInline, deployed as a physical appliance on-premises
   Reveal Solution Discussion Select an answer first