Skip to content
aiqbench takeAI Products

aiqbench analysis

This empirical study provides a longitudinal analysis of AI coding agents within real-world software development workflows. It offers critical insights into how agentic pull requests compare to human-generated ones regarding merge rates, task distribution evolution, and overall code quality implications over time. For practitioners and researchers tracking the integration of LLMs in engineering pipelines, this data clarifies both current capabilities and emerging limitations as these tools mature across different development stages.
Sources coveringarXiv, Fortune, AI Alignment Forum, LessWrong
What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

How Do AI Coding Agents Contribute to Software Development? an Empirical Study of Agentic Pull Requests

arXiv3 min read

arXiv:2607.21832v1 Announce Type: cross Abstract: Recent advances in large language models and their rapid adoption across software engineering tasks have made Artificial Intelligence (AI) coding agents an integral component of modern software development workflows. While developers increasingly benefit from these coding agents, their impact on software quality remains insufficiently understood. In particular, how agentic contributions evolve across the software development lifecycle has not been thoroughly investigated. This study aims to characterize agentic pull requests (PR) in comparison to human generated PRs and to examine how their properties change across different stages of the development lifecycle. Using the AIDev dataset, we first analyze how differences in merge rates between agentic and human generated PRs vary over time. We then identify the types of development tasks where AI coding agents are predominantly applied and investigate how these task distributions evolve across development quarters. Finally, we compare a set of key characteristics of agentic and human generated PRs, focusing on their implications for software quality and their temporal dynamics. Overall, our findings provide an empirical and longitudinal perspective on the role of AI coding agents in software development, offering a more nuanced understanding of their benefits and limitations in real-world practices.

Title:How Do AI Coding Agents Contribute to Software Development? an Empirical Study of Agentic Pull Requests

Abstract:Recent advances in large language models and their rapid adoption across software engineering tasks have made Artificial Intelligence (AI) coding agents an integral component of modern software development workflows. While developers increasingly benefit from these coding agents, their impact on software quality remains insufficiently understood. In particular, how agentic contributions evolve across the software development lifecycle has not been thoroughly investigated. This study aims to characterize agentic pull requests (PR) in comparison to human generated PRs and to examine how their properties change across different stages of the development lifecycle. Using the AIDev dataset, we first analyze how differences in merge rates between agentic and human generated PRs vary over time. We then identify the types of development tasks where AI coding agents are predominantly applied and investigate how these task distributions evolve across development quarters. Finally, we compare a set of key characteristics of agentic and human generated PRs, focusing on their implications for software quality and their temporal dynamics. Overall, our findings provide an empirical and longitudinal perspective on the role of AI coding agents in software development, offering a more nuanced understanding of their benefits and limitations in real-world practices.

Software Engineering (cs.SE); Machine Learning (cs.LG)

arXiv:2607.21832 [cs.SE]

(or arXiv:2607.21832v1 [cs.SE] for this version)

https://doi.org/10.48550/arXiv.2607.21832 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Full text under the publisher's open licence · 306 words

Related coverage