AI blocks security features, leaving vibe coders exposed
A Reddit user posted a screenshot raising a concern: when someone tries to write actual security features into their app (like login or ), AI models such as flag or refuse the request as risky. Meanwhile, people or with recognized security clearance reportedly don't hit the same restrictions. The worry is that so-called '' — people who rely on AI to write code rather than writing it themselves — can't get the AI's help to properly secure their apps, which could leave those apps with weak or missing security once they're shipped.
Key points
- Claim: AI models flag/refuse coding requests related to security features as risky
- Users with security clearance reportedly face fewer such restrictions
- '' who can't write all the code themselves have no easy workaround
- Concern: this could result in apps shipping with weak security