Security Audit Reveals Thousands of Lovable-Built Apps Expose Sensitive User Data

As generative AI lowers the barrier to full-stack software development, a comprehensive security audit of 159,041 projects built on the Lovable platform has uncovered significant vulnerabilities. Researchers found numerous applications exposing confidential user records, passport details, and private communications without requiring authentication.

Tracing its roots back to open-source origins, Lovable enables users to spin up sophisticated web applications from natural language prompts. However, the findings underscore a critical industry challenge: while AI excels at generating functional frontend and backend code, enforcing rigorous database security and access control remains a vital human responsibility.