Generative AI Enables Software Engineering, with Adoption Now Widely Integrated into Daily Tasks

The rapid integration of generative AI tools into software engineering is transforming how code is written, tested and maintained, yet a clear understanding of this adoption remains elusive. Görkem Giray from Eindhoven University of Technology and Izmir Institute of Technology, along with Onur Demirörs from Izmir Institute of Technology and Marcos Kalinowski from Pontifical Catholic University of Rio de Janeiro, alongside Daniel Mendez et al., now present compelling empirical evidence detailing the current landscape of generative AI in practice. Their work reveals widespread use of these tools across various software engineering tasks, including implementation, verification and validation, and personal assistance, with practitioners reporting significant benefits such as reduced development time and improved code quality. While acknowledging challenges related to output reliability, security and the need for robust validation, the research demonstrates that generative AI is becoming deeply embedded in daily workflows and is anticipated to reshape, rather than replace, the roles of software engineers.

In software engineering, generative artificial intelligence (GenAI) offers promising support for several activities. Despite increasing adoption, empirical evidence regarding how GenAI is used in practice, the benefits it provides, the challenges it introduces, and its broader organizational and societal implications remains limited. This study aims to provide an overview of the status of GenAI adoption in software engineering. It investigates the status of GenAI adoption, associated benefits and challenges, the institutionalization of tools and techniques, and anticipated long-term impacts on software engineering professionals and the community. The research team conducted an internationally distributed questionnaire-based survey to collect insights from software engineering practitioners.

Global Survey Maps Generative AI Adoption

The study pioneers a rigorous, internationally distributed investigation into the adoption of Generative AI within Software Engineering, employing a questionnaire-based survey to gather detailed insights from practitioners across the globe. Researchers designed a survey instrument combining both closed-ended and open-ended questions, enabling the collection of quantifiable trends alongside nuanced qualitative data, ultimately receiving 204 responses representing 37 countries. This approach allows for a comprehensive understanding of how GenAI tools are currently integrated into daily workflows and the perceived impacts of their use., To ensure the robustness of findings, the team meticulously analyzed qualitative responses through systematic coding and categorization, identifying recurring themes and patterns in practitioner experiences. Recognizing the limitations inherent in random sampling, scientists implemented bootstrapping techniques, a statistical method for estimating confidence intervals, and conservatively reported these intervals to account for potential sampling bias.

This careful statistical treatment strengthens the validity of the observed trends and provides a more reliable assessment of GenAI’s influence., The research highlights widespread integration of GenAI tools, particularly in areas such as implementation, verification, validation, personal assistance, and maintenance, demonstrating a deep embedding within existing software engineering practices. Practitioners consistently reported benefits including reductions in cycle time, improvements in software quality, enhanced knowledge work support, and overall productivity gains, although objective measurement of these gains remains an area for future investigation. The study’s methodology, therefore, not only documents current adoption levels but also establishes a foundation for future research focused on quantifying the impact of GenAI on software engineering performance.

Generative AI Transforms Software Engineering Practice

The research team conducted an internationally distributed survey, gathering responses from 204 software engineering practitioners across 37 countries, to investigate the adoption of generative AI tools. Results demonstrate widespread integration of these tools into daily work, particularly for implementation, verification, validation, personal assistance, and maintenance tasks, indicating a significant shift in how software is developed and maintained. Practitioners consistently reported substantial benefits from using GenAI, including reductions in cycle time and improvements in software quality, alongside enhanced support for knowledge work and overall productivity gains., While objective measurement of these productivity and quality improvements remains limited, the study highlights a clear perception of value among users. Experiments revealed that practitioners are actively institutionalizing these tools and techniques, though access to tools is prioritized more than comprehensive training or governance structures.

The team also recorded significant challenges, including instances of incorrect or unreliable outputs, difficulties with prompt engineering, and the overhead associated with validating generated code. Security and privacy concerns, as well as the risk of overreliance on these systems, also emerged as key obstacles to broader adoption., Practitioners anticipate that GenAI will redefine their roles rather than replace them, yet express moderate concern regarding potential contraction in the job market and the need for evolving skill sets. The findings corroborate growing attention within the software engineering community, delivering perceived value while simultaneously introducing new technical, organizational, and socio-technical challenges that require careful consideration. This work establishes a baseline understanding of GenAI adoption, paving the way for more systematic approaches to ensure its successful, sustainable, and responsible integration into software engineering practices.

Generative AI Transforms Software Engineering Practice

This research demonstrates widespread adoption of generative AI tools within the software engineering community, with practitioners integrating these technologies into daily tasks spanning implementation, verification, validation, and maintenance. The study highlights perceived benefits including reduced development cycles, improved software quality, and gains in productivity, particularly in knowledge-based work. While objective measurement of these improvements remains limited, the data clearly indicates a positive reception and increasing reliance on these new tools., However, the team also identified significant challenges accompanying this adoption, notably the potential for inaccurate or unreliable outputs, the complexities of prompt engineering, and concerns surrounding security and privacy. Practitioners acknowledge the need for careful validation of generated content and express awareness of the risks associated with over-dependence on these systems.

The research reveals that organizations commonly prioritize tool access, but often lack comprehensive training programs and governance structures to support responsible implementation. Participants anticipate that generative AI will reshape software engineering roles, though moderate concerns exist regarding potential impacts on the job market and required skillsets., The authors acknowledge that further investigation is needed to establish robust metrics for evaluating productivity and quality gains, and to fully understand the long-term socio-technical implications of these technologies. Future work should focus on developing systematic approaches to integration, moving beyond ad-hoc adoption to ensure sustainable and responsible use of generative AI in software engineering practices.