This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
Choose the best response for each question.
What is one of the main challenges of securing AI systems compared to traditional software?
The nondeterministic nature of generative AI systems
AI is fundamentally insecure and can't be protected
AI is too new to have any security frameworks
Which jailbreak technique involves sending an AI model a series of progressively escalating prompts across multiple conversation turns?
DAN (Do Anything Now)
Encoding attack
Crescendo
Which is the middle layer of the three AI architecture layers?
AI Usage layer
AI Platform layer
AI Application layer
What makes indirect prompt injection (XPIA) dangerous compared to direct prompt injection?
It requires more technical skill to execute
The malicious instructions are hidden in content the AI retrieves, not in the user's prompt
It can only be performed by authorized system administrators
Which of the following is a risk associated with AI overreliance?
Users accept AI-generated output without verification, leading to errors in critical decisions
AI systems become slower over time due to excessive use
AI models are unable to process complex queries
You must answer all questions before checking your work.
Was this page helpful?
Need help with this topic?
Want to try using Ask Learn to clarify or guide you through this topic?