The Impact of Generative Artificial Intelligence on Software Development Teams During the Construction Phase: Systematic Mapping Study

Abstract

During the construction phase of software development, developers often perform repetitive coding tasks, such as creating boilerplate structures. Automating these tasks can significantly improve coding efficiency. Generative Artificial Intelligence (GenAI) has emerged as a valuable tool in this context, assisting developers by suggesting code snippets or generating functions from natural language prompts, thereby accelerating the development process. However, the use of GenAI also introduces certain risks, particularly in terms of security and the occurrence of “hallucination”, instances where the AI suggests non-existent functions or features, potentially leading to developer frustration. This Systematic Mapping Study aims to examine the current state of GenAI adoption among software development teams during the construction phase. The study analyzes both the benefits and risks associated with its use, its impact on productivity during construction activities, and offers recommendations for its effective integration. To conduct this review, we consulted four academic databases and included studies published between 2020 and 2025. After applying inclusion and exclusion criteria, we identified 44 primary studies. Our analysis reveals that integrating GenAI can substantially enhance development efficiency. Nonetheless, its use also entails risks that must be carefully managed to avoid negative consequences for both the development team and the resulting software. Furthermore, the findings highlight the positive influence of GenAI on the productivity of specific construction-related tasks. In light of these insights, we recommend that development teams adopt targeted strategies to maximize the benefits of GenAI while mitigating the associated risks.

Ähnliche Arbeiten