Secure AI SDLC

1. What is the main goal of incorporating security into the AI Software Development Life Cycle (SDLC)?

To avoid code reviews To automate security checks in the development pipeline To ensure developers can ignore security practices To simplify the deployment process

2. Which scanning process should be performed on every build to catch vulnerabilities like injection sinks and unsafe deserialization?

Dynamic analysis Static analysis (SAST) Manual code review Penetration testing

3. What principle should be followed regarding database logins in a secure AI SDLC?

They should have admin rights for all operations They should be able to perform any database operation They should follow least privilege practices They should be hardcoded in the application code

4. Why is it essential for guardrails in an AI system to have regression tests?

To make them easier to debug To ensure they are properly documented To protect against accidental removal during refactoring To increase performance

Continue: Why AI Governance → Review this lesson Retake quiz