Experts point out that having AI write too much code causes inspection costs to explode, making it important to reduce the amount generated

WaveMaker cofounder and CTO Deepak Anapalli argued in InfoWorld that AI-generated code has limits if teams rely only on post-generation inspection.

He says developers should shift to an “AI assembly model” that minimizes new code by reusing existing, verified components.

The goal is to cut review, fixing, maintenance, and audit costs as AI code output keeps growing.

The article underscores that efficiency is not just about generating faster, but about managing total quality and security cost.