"TODO: Fix the Mess Gemini Created": Towards Understanding GenAI-Induced Self-Admitted Technical Debt

TL;DR

This paper investigates how developers acknowledge technical debt in AI-assisted code comments, revealing patterns of self-admitted issues related to testing, adaptation, and understanding in GenAI-influenced software development.

Contribution

It introduces the concept of GIST, a new lens to understand self-admitted technical debt caused by generative AI in code comments, based on empirical analysis.

Findings

Developers frequently mention testing postponement and incomplete adaptation.

AI assistance influences the timing and reasons for technical debt.

81 instances of self-admitted technical debt were identified from 6,540 comments.

Abstract

As large language models (LLMs) such as ChatGPT, Copilot, Claude, and Gemini become integrated into software development workflows, developers increasingly leave traces of AI involvement in their code comments. Among these, some comments explicitly acknowledge both the use of generative AI and the presence of technical shortcomings. Analyzing 6,540 LLM-referencing code comments from public Python and JavaScript-based GitHub repositories (November 2022-July 2025), we identified 81 that also self-admit technical debt(SATD). Developers most often describe postponed testing, incomplete adaptation, and limited understanding of AI-generated code, suggesting that AI assistance affects both when and why technical debt emerges. We term GenAI-Induced Self-admitted Technical debt (GIST) as a proposed conceptual lens to describe recurring cases where developers incorporate AI-generated code while explicitly expressing uncertainty about its behavior or correctness.