Study finds 434 exploitable flaws in AI-generated apps

Security analysis of vibe-coded apps revealed 434 exploitable flaws, with common issues including denial-of-service, authorization bypass, and secrets exposure. The findings highlight security risks in AI-generated code.