On the Adoption of AI Coding Agents in Open-source Android and iOS Development
Muhammad Ahmad Khan, Hasnain Ali, Muneeb Rana, Muhammad Saqib Ilyas, Abdul Ali Bangash
TL;DR
The study investigates how AI coding agents contribute to open-source mobile development by analyzing 2,901 AI-authored PRs across 193 Android and iOS projects from the AIDev dataset. It develops a 13-category task taxonomy using GPT-5 and validates it, then assesses PR acceptance and resolution across platforms, agents, and task categories. Key findings show Android exhibiting higher acceptance and more agent-level variation than iOS, with routine tasks like localization, UI, and fixes being most readily accepted, while structural changes are slower to resolve. The results offer baseline, platform-aware guidance for integrating agentic coding in mobile workflows and inform the design of future platform-specific AI coding assistants.
Abstract
AI coding agents are increasingly contributing to software development, yet their impact on mobile development has received little empirical attention. In this paper, we present the first category-level empirical study of agent-generated code in open-source mobile app projects. We analyzed PR acceptance behaviors across mobile platforms, agents, and task categories using 2,901 AI-authored pull requests (PRs) in 193 verified Android and iOS open-source GitHub repositories in the AIDev dataset. We find that Android projects have received 2x more AI-authored PRs and have achieved higher PR acceptance rate (71%) than iOS (63%), with significant agent-level variation on Android. Across task categories, PRs with routine tasks (feature, fix, and ui) achieve the highest acceptance, while structural changes like refactor and build achieve lower success and longer resolution times. Furthermore, our evolution analysis shows improvement in PR resolution time on Android through mid-2025 before it declined again. Our findings offer the first evidence-based characterization of AI agents effects on OSS mobile projects and establish empirical baselines for evaluating agent-generated contributions to design platform aware agentic systems.