To What Extent Does Agent-generated Code Require Maintenance? An Empirical Study

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Sawada, Shota, Shirai, Tatsuya, Kashiwa, Yutaro, Yamaguchi, Ken'ichi, Iwata, Hiroshi, Iida, Hajimu
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910204552019968
author Sawada, Shota
Shirai, Tatsuya
Kashiwa, Yutaro
Yamaguchi, Ken'ichi
Iwata, Hiroshi
Iida, Hajimu
author_facet Sawada, Shota
Shirai, Tatsuya
Kashiwa, Yutaro
Yamaguchi, Ken'ichi
Iwata, Hiroshi
Iida, Hajimu
contents LLM-based autonomous coding agents have reshaped software development. While these agents excel at code generation, open questions persist about the long-term maintainability of AI-generated code. This study empirically investigates the maintenance extent, human involvement, and modification types of AI-generated files versus human-authored code. Using the AIDev dataset of AI-generated pull requests and GitHub, we analyzed over 1,000 files and approximately 3,200 changes from 100 popular repositories. Our findings show that: (i) AI-generated files receive less frequent maintenance than human-authored code, with updates affecting only a small fraction of file size; (ii) the most frequent modifications to AI code are feature extensions, whereas human updates focus on bug fixes, and (iii) human developers perform the large majority of this maintenance.
format Preprint
id arxiv_https___arxiv_org_abs_2605_06464
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle To What Extent Does Agent-generated Code Require Maintenance? An Empirical Study
Sawada, Shota
Shirai, Tatsuya
Kashiwa, Yutaro
Yamaguchi, Ken'ichi
Iwata, Hiroshi
Iida, Hajimu
Software Engineering
LLM-based autonomous coding agents have reshaped software development. While these agents excel at code generation, open questions persist about the long-term maintainability of AI-generated code. This study empirically investigates the maintenance extent, human involvement, and modification types of AI-generated files versus human-authored code. Using the AIDev dataset of AI-generated pull requests and GitHub, we analyzed over 1,000 files and approximately 3,200 changes from 100 popular repositories. Our findings show that: (i) AI-generated files receive less frequent maintenance than human-authored code, with updates affecting only a small fraction of file size; (ii) the most frequent modifications to AI code are feature extensions, whereas human updates focus on bug fixes, and (iii) human developers perform the large majority of this maintenance.
title To What Extent Does Agent-generated Code Require Maintenance? An Empirical Study
topic Software Engineering
url https://arxiv.org/abs/2605.06464