Should AI-generated code be reviewed differently in code review?

Yes — AI-generated code warrants extra scrutiny in a few areas: security vulnerabilities (AI frequently misses auth checks or SQL injection risks), subtle logic errors that look correct at a glance, and hallucinated API calls. Many teams add a comment tagging AI-assisted sections so reviewers know to look more carefully. Tools like CodeRabbit and Bito can also automatically review PRs for AI-introduced issues.