Type: url
Original: https://arxiv.org/pdf/2512.23327
Status: generated
Summary
“An Empirical Study of Generative AI Adoption in Software Engineering” (Giray, Demirörs, Kalinowski and Mendez; arXiv:2512.23327) reports an international questionnaire survey of 204 software engineering practitioners from 37 countries, collected between May and November 2025, on how GenAI tools are used in practice, the benefits and challenges they bring, how organizations institutionalize them, and what practitioners expect their long-term impact to be. About 80% of respondents use GenAI tools, mostly for implementation, followed by verification and validation, personal assistance and maintenance, and they report reduced cycle time, perceived quality improvements and productivity gains — yet 58% use no objective metric to measure productivity or quality. Organizations mostly provide tool access rather than training, policies or KPIs, and practitioners largely expect GenAI to redefine rather than replace their roles while anticipating a contraction of the job market.