Integrating GenAI into Software Work: Product Quality, Collaboration, and Process Impacts under an ISO/IEC 25010 Lens
Abstract
Context: Generative Artificial Intelligence (GenAI), particularly Large Language Models (LLMs), has rapidly emerged as a transformative force in software engineering, supporting activities such as code generation, refactoring, testing, and documentation. Despite their growing adoption, empirical evidence remains limited regarding how LLM-based tools influence established dimensions of software product quality such as those defined in ISO/IEC 25010 as well as collaboration and innovation within software teams. Goal: This study investigates how the integration of GenAI tools affects software quality, human-centered factors, and development processes across the software lifecycle. We aim to provide a holistic understanding of both the opportunities and risks associated with their adoption in professional environments. Method: We conducted a survey with 121 software practitioners from 26 Brazilian states, spanning public and private sectors. The questionnaire comprised 38 closed- and open-ended questions. Quantitative data were analyzed using descriptive statistics analysis, while qualitative answers underwent open and axial coding following Grounded Theory principles. Results: Practitioners reported positive perceptions of GenAI in functional suitability, performance efficiency, maintainability, and flexibility, as well as in usability and adaptability. They emphasized productivity gains, faster bug detection, and automation of repetitive tasks, but expressed concerns about reliability, security, and safety. Open-ended responses revealed improvements in collaboration, communication, and innovation, particularly through rapid prototyping and creative ideation, alongside risks such as overreliance, diminished peer interaction, and a lack of long-term evidence on software quality. Conclusions: GenAI tools are perceived as important complements to software development, enhancing productivity, code quality, and team collaboration when used under human-in-the-loop oversight. Their integration demands governance mechanisms that ensure reliability, accountability, and sustainable quality improvements.
Full text
Q.36. How has the adoption of generative AI tools influenced your software development practices, and in what ways has this affected the quality of the software you produce? Productivity & Efficiency Category Sub-category # Quotes Practitioners's Answer Respondent Increases productivity e delivery speed 25 2 AI speeds up achieving results. 6 AI accelerates development and learning. 7 14 17 AI speeds reviews and improves usability by catching overlooked issues. 17 AI boosts productivity, quality, and innovation. 22 AI suggestions accelerate productivity. 24 25 27 AI reduces time in tasks. Time reduction. 31 AI improves speed, quality, and comprehension. Speed, quality, and understanding. 33 AI accelerates code production. Code production with more speed. 34 AI accelerates processes. Accelerating processes. 35 48 57 AI accelerates both coding and skill acquisition. 58 65 68 AI’s main advantage is faster delivery. 72 AI accelerates development and learning. 73 AI increases productivity. AI helps us improve productivity. 80 86 98 108 Automation of repetitive tasks 15 AI reduces repetitive tasks. 4 5 10 AI automates repetitive tasks. Automates repetitive tasks. 16 21 49 AI took over repetitive functions. 50 52 56 61 AI automates routine and repetitive work. It automates repetitive tasks. 67 AI reduces manual effort in repetitive activities. 70 AI supports repetitive tasks and error detection. 71 91 99 Code generation support 1 AI helps generate boilerplate and simple snippets. 39 Cognitive load reduction 1 118 9 20 41 AI helps in error identification. 43 46 49 60 AI accelerates delivery with improved quality and fe Time agility, allowing faster and better deliv The main factor is delivery speed. With AI y Greater speed in development and learning AI speeds up early development by providing base It facilitated the start of the development cy AI accelerates coding, improves documentation and They help generate code quickly, saving tim Increase in productivity, improvement in cod With AI introduction, its suggestions make m AI boosts efficiency, error detection, and standardiz The adoption of generative AI tools in softw AI enhances productivity, quality, prototyping, and i Increased productivity and software quality, AI accelerates development but increases code wo It sped things up, but now the code has mo AI ensures faster delivery while keeping results con The main factor is delivery speed. With AI y Greater speed in development and learning AI accelerates early development by offering base It facilitated the beginning of the developme AI accelerates coding, improves testing, and enhan They help generate code quickly, saving tim The greatest benefit of AI tools is enabling Greater speed of development and learning AI enhanced both productivity and software quality. Generative AI tools increased the productiv AI boosts productivity and aligns deliveries with cus The use of generative AI has boosted prod AI speeds up delivery while maintaining quality, boo Generative AI increased delivery speed wit The use of AI tools minimized the execution AI helps with repetitive tasks and error detection, b It helps a lot in repetitive tasks and identifyi Improvement in documentation and recommendatio Repetitive tasks were accelerated with the AI handles repetitive/low-complexity tasks, reducing It provided a broader view, being assertive Copilot standardized code, reduced cognitive load, With Github Copilot, the team made code m Some functions that were previously done r AI automates tasks, accelerates development, and Generative AI automated repetitive tasks, s AI helps with repetitive tasks and error detection, b It helps a lot with repetitive tasks and identi AI speeds up repetitive tasks and supports docume Repetitive tasks were accelerated with the It minimized the execution of repetitive task AI tools help a lot with repetitive tasks and AI automates repetitive tasks, freeing focus for com The adoption of generative AI tools has stre AI speeds up boilerplate code generation, letting te Generative AI sped up writing boilerplate co I use it to generate boilerplate and some sh AI reduced cognitive load by providing helpful sugg Generative AI tools helped reduce the team AI boosts productivity in error detection and correct It increased productivity in correcting and id AI anticipates possible errors, enhancing performan The tool has a certain predictability regardin AI assists with functions, code, and bug fixes, but re Generative AI works as an assistant. It facil I can identify errors in the programmed cod Limited use of AI: mainly for clarifying doubts and e I cannot say how LLM influenced developm Copilot standardized code, reduced cognitive load, With Github Copilot, the team made code m AI enhances productivity by accelerating bug detec It increased productivity in correcting and id
Software Quality & Maintenance Testing & Documentation Limitations / Risks Error detection and correction 15 68 75 AI accelerates error identification. 77 84 AI reduced coding errors. The tools reduced code errors 89 92 95 101 Code quality and best practices 12 AI supports writing more effective code. 8 AI encourages good coding practices and quality. 42 47 51 59 AI improved overall team code quality. 74 76 AI improves code quality in products. 82 83 AI influenced both coding and testing practices. 88 AI enables higher-quality solutions. 90 68 Code refactoring and sustainability 1 103 Legacy code analysis and modernization 1 117 Support in testing and documentation 12 AI supports tests, documentation, improvements, and reducing technical debt. 11 AI speeds reviews and improves usability by catching overlooked issues. 17 23 47 61 62 66 68 AI supports contextualized documentation. 81 93 97 102 Testing support and code review 3 15 106 107 Documentation support 2 10 29 Necessity of human validation 6 Dependency on user’s prior logic 3 13 41 55 64 Overreliance on AI 3 Reliance on AI outputs. Trusting what AI provides. 12 25 Trusting what AI provides. 63 Inadequate or incomplete solutions 2 1 48 18 AI accelerates coding, improves testing, and enhan They help generate code quickly, saving tim AI acts as an assistant for code improvement and b Generative AI works as an assistant, suppo With AI tools I can quickly identify errors in AI changed how we approach documentation and b Generative AI tools changed our view on ho AI enhances early bug detection and fixing, leading Generative AI tools have enhanced our abi AI reduces human errors in coding, increasing prod With generative AI tools, we have reduced AI increases agility in bug fixing, reducing resolutio The adoption of generative AI brought grea Generative AI helped me in software develo AI tools are important for good coding pract AI improves code formatting, structure, and context It started generating better formatted and s AI improves code structure and promotes best prac The main improvement in quality was supp AI contributes to more efficient and effective code w Generative AI helped me in software develo The use of AI tools allowed us to improve th AI encourages best practices and enhances quality AI tools are important for good coding pract AI tools help us improve the quality of the c AI reshaped coding practices, leading to better cod Generative AI tools influenced our way of g It influenced our way of producing code and The use of LLMs allows us to deliver better AI accelerates coding, improves testing, and enhan They help generate code quickly, saving tim AI improves code refactoring, creating cleaner and The use of generative AI increased efficien AI enabled efficient analysis of large legacy codeba With generative AI, we were able to analyze AI aids code comprehension, standardization, docu Helps in understanding code written by othe AI improves code formatting, structure, and context It started generating better formatted and s AI speeds up repetitive tasks and supports docume Repetitive tasks were accelerated with the AI assists in testing, documentation, and reducing t Creation of unit tests, Java documentation, AI helps in testing and reviews, but its outputs often I mainly use LLMs to generate tests and pe AI accelerates coding, improves testing, and enhan They help generate code quickly, saving tim With AI tools we started generating docume AI improves productivity through code suggestions By using generative AI for code suggestion AI-generated automated tests expand coverage an Generative AI support in writing automated AI generates clearer and more contextual documen Generative AI supported us in creating clea AI helps with tests and code review, but results ofte I primarily use LLMs to generate tests and AI strengthens regression test automation, improvin Generative AI tools strengthened regressio AI increases confidence in test coverage, reducing The use of generative AI gave us greater co AI accelerates repetitive tasks and improves docum Repetitive tasks were accelerated with the AI improves speed in documentation and content c Improved response time in producing conte Nao afetou em nada, as ferramentas de IA Code must be reviewed against project design patt Always review the code based on the proje AI assists with functions, code, and bug fixes, but re Generative AI works as an assistant. It facil AI explains its code review suggestions, requiring h AI not only reviews the code but also expla AI outputs must be validated against design pattern Always review the code based on the proje AI boosts efficiency, error detection, and standardiz The adoption of generative AI tools in softw Some developers rely heavily on AI-generated outp AI generates shorter code, but not always effective Sometimes the code is shortened, but the s AI accelerates development but increases code wo It sped things up, but now the code has mo AI speeds reviews and improves usability by catchi I can do reviews faster, and usability is imp
Collaboration & Processes Innovation & Learning Requirements Technical Debt Project Management Compliance Security Usability Code review and error handling 6 AI automates code review tasks. 36 55 69 96 Code standardization 2 100 49 Team integration 1 105 Stakeholder communication and prototyping 1 115 Innovation and prototyping 4 53 94 104 114 Learning and skills acquisition 3 19 113 121 Internationalization and localization 1 111 Accessibility support 1 112 Requirements analysis and elicitation 1 109 Requirements traceability 1 119 Technical debt reduction 1 62 Project management and planning support 1 110 Regulatory compliance and standards alignment 1 116 Security and vulnerability management 1 120 Usability improvement 18 The use of AI contributed to automating tas AI explains its code review suggestions, requiring h AI not only reviews the code but also expla AI makes reviews faster and improves usability by s I can do reviews faster, and usability improv AI makes code review faster and more effective, co The adoption of generative AI has improved AI standardizes code style, improving readability an Generative AI tools helped standardize cod Copilot standardized code, reduced cognitive load, With Github Copilot, the team made code m AI reduces onboarding time, helping new develope Generative AI support reduced the onboard AI improved stakeholder communication through qu The use of generative AI improved commun AI expands solution space with alternative code sug With AI generating alternatives and code su AI accelerates prototyping, enabling faster validatio Generative AI has helped us accelerate pro AI supports exploring design alternatives, fostering Generative AI improved our ability to explor AI stimulated innovation by suggesting alternative s Generative AI stimulated innovation by prop AI helps when lacking expertise in a programming l In code solutions where I did not master the AI reduces the learning curve for new languages an Generative AI support was essential in redu AI enhanced continuous learning by offering contex Generative AI tools enhanced our capacity AI supports software internationalization with quick Generative AI tools contributed to software AI improves software accessibility by suggesting ad Generative AI helped us improve software a AI expands requirement reviews, suggesting uncon Generative AI expanded our ability to review AI improved traceability between requirements and Generative AI improved traceability betwee AI assists in testing, documentation, and reducing t Creation of unit tests, Java documentation, AI provides insights during sprint planning, aiding e The use of generative AI provided valuable AI supported regulatory compliance by suggesting Generative AI supported regulatory complia AI identified vulnerability patterns early, strengtheni The use of generative AI helped us identify AI speeds reviews and improves usability by catchi I can do reviews faster, and usability is imp