Every AI Commit Is Someone's Future Legacy Code

Short summary

AI coding tools are simultaneously helping enterprises modernize decades-old legacy systems and creating a new wave of AI-generated legacy code that developers avoid maintaining. GitClear research shows legacy refactoring dropped 74% since 2023, while Google's DORA report links every 25% increase in AI usage to 7.2% more system instability. The honest answer is that AI helps at the expensive enterprise end but quietly accelerates tech debt at the everyday developer level.

• •AI helps enterprises modernize old COBOL/RPG/IBM i systems but creates new legacy code at the daily developer level
• •Legacy refactoring dropped 74% since 2023 as developers prefer building new code over cleaning up existing code
• •Google DORA report: every 25% increase in AI usage adds 7.2% more system instability

Generated with AI, which can make mistakes.