Q50: Why is human review of AI-generated security code essential before production deployment?
Generative AI models predict code patterns from training data, which may contain legacy security vulnerabilities (e.g., SQL injection, hardcoded secrets, weak encryption). Human oversight ensures compliance with security standards.
Key Takeaways
- AI suggestions are not guaranteed to be secure.
- Always validate security-sensitive code (auth, encryption, DB queries).
Knowledge Check
Why must developers review AI-generated code for security vulnerabilities?