Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Navigating the Complexity of Generative AI Adoption in Software Engineering

Daniel Russo

TL;DR

This study investigates why software engineers adopt Generative AI tools by integrating TAM, DOI, and SCT into a new HACAF model. It combines a qualitative Gioia-based inquiry (n=100) with a quantitative PLS-SEM validation (n=183), revealing that adoption in early stages is primarily driven by compatibility with existing workflows rather than perceived usefulness, social influence, or personal innovativeness. The HACAF model shows that perceptions about the technology influence compatibility, which in turn drives intention to use, while social factors and personal/environmental factors have weaker direct effects; IPMA highlights compatibility and perceived usefulness as key levers. These findings inform tool design and organizational strategies, emphasizing seamless integration, practical utility, and user-centered, human-in-the-loop approaches to accelerate responsible AI adoption in software engineering. The work also underscores the need for longitudinal studies to track evolving adoption dynamics as Generative AI tools mature.

Abstract

In this paper, the adoption patterns of Generative Artificial Intelligence (AI) tools within software engineering are investigated. Influencing factors at the individual, technological, and societal levels are analyzed using a mixed-methods approach for an extensive comprehension of AI adoption. An initial structured interview was conducted with 100 software engineers, employing the Technology Acceptance Model (TAM), the Diffusion of Innovations theory (DOI), and the Social Cognitive Theory (SCT) as guiding theories. A theoretical model named the Human-AI Collaboration and Adaptation Framework (HACAF) was deduced using the Gioia Methodology, characterizing AI adoption in software engineering. This model's validity was subsequently tested through Partial Least Squares - Structural Equation Modeling (PLS-SEM), using data collected from 183 software professionals. The results indicate that the adoption of AI tools in these early integration stages is primarily driven by their compatibility with existing development workflows. This finding counters the traditional theories of technology acceptance. Contrary to expectations, the influence of perceived usefulness, social aspects, and personal innovativeness on adoption appeared to be less significant. This paper yields significant insights for the design of future AI tools and supplies a structure for devising effective strategies for organizational implementation.

Navigating the Complexity of Generative AI Adoption in Software Engineering

TL;DR

This study investigates why software engineers adopt Generative AI tools by integrating TAM, DOI, and SCT into a new HACAF model. It combines a qualitative Gioia-based inquiry (n=100) with a quantitative PLS-SEM validation (n=183), revealing that adoption in early stages is primarily driven by compatibility with existing workflows rather than perceived usefulness, social influence, or personal innovativeness. The HACAF model shows that perceptions about the technology influence compatibility, which in turn drives intention to use, while social factors and personal/environmental factors have weaker direct effects; IPMA highlights compatibility and perceived usefulness as key levers. These findings inform tool design and organizational strategies, emphasizing seamless integration, practical utility, and user-centered, human-in-the-loop approaches to accelerate responsible AI adoption in software engineering. The work also underscores the need for longitudinal studies to track evolving adoption dynamics as Generative AI tools mature.

Abstract

In this paper, the adoption patterns of Generative Artificial Intelligence (AI) tools within software engineering are investigated. Influencing factors at the individual, technological, and societal levels are analyzed using a mixed-methods approach for an extensive comprehension of AI adoption. An initial structured interview was conducted with 100 software engineers, employing the Technology Acceptance Model (TAM), the Diffusion of Innovations theory (DOI), and the Social Cognitive Theory (SCT) as guiding theories. A theoretical model named the Human-AI Collaboration and Adaptation Framework (HACAF) was deduced using the Gioia Methodology, characterizing AI adoption in software engineering. This model's validity was subsequently tested through Partial Least Squares - Structural Equation Modeling (PLS-SEM), using data collected from 183 software professionals. The results indicate that the adoption of AI tools in these early integration stages is primarily driven by their compatibility with existing development workflows. This finding counters the traditional theories of technology acceptance. Contrary to expectations, the influence of perceived usefulness, social aspects, and personal innovativeness on adoption appeared to be less significant. This paper yields significant insights for the design of future AI tools and supplies a structure for devising effective strategies for organizational implementation.
Paper Structure (96 sections, 4 figures, 20 tables)

This paper contains 96 sections, 4 figures, 20 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Assessing Generated Code: Correctness and Quality
  4. Evaluation Criteria: Diverse Approaches
  5. Enhancing Code Productivity
  6. Comparing Methods
  7. Pedagogical Concerns
  8. AI's Influence on Software Development Process
  9. Community Influence and Trust in AI Tools
  10. Generative AI in Non-Coding Activities
  11. Usability of AI Programming Assistants
  12. Theory Generation
  13. Theoretical foundation
  14. Technology Acceptance Model (TAM)
  15. Diffusion of Innovation Theory (DOI)
  16. ...and 81 more sections

Figures (4)

  • Figure 1: The Human-AI Collaboration and Adaptation Framework (HACAF)
  • Figure 2: HACAF's structural model with $R^2$ and path coefficients (*** p<0.001, (NS) p>0.05).
  • Figure 3: Importance-Performance Map Analysis of Intention to Use.
  • Figure 4: Residuals vs. predicted values for the 'Intention to Use' construct. The residuals are mostly randomly scattered around zero, suggesting that the model's errors are random and that the model is correctly specified.