AI assisted requirements engineering for system development

Objective: To analyze the relationship between AI-assisted requirements engineering and the development of intelligent systems, considering its contribution to the elicitation, drafting, analysis, classification, validation and traceability of software requirements. Methodology: A quantitative, non-experimental, cross-sectional and descriptive-correlational research was carried out with a sample of 143 students, teachers and professionals linked to software development, information technologies, functional analysis and systems engineering. The information was collected using a 24-item questionnaire with a Likert-type scale, validated by specialists and with a Cronbach's alpha coefficient of 0.87. For the analysis, descriptive statistics and Spearman's correlation coefficient were applied. Results: A favorable perception was identified about the use of artificial intelligence in requirements engineering activities. The main contributions were observed in the drafting, classification, clarity, coherence, precision and reduction of ambiguities of the requirements. The correlational analysis evidenced a positive and statistically significant relationship between AI-assisted requirements engineering and the development of intelligent systems, with the quality of requirements being highlighted as the dimension with the greatest relationship. Limitations associated with bias, misinterpretation, technological dependence, and lack of contextualization were also recognized. Conclusions: Artificial intelligence can strengthen the efficiency and quality of requirements engineering, facilitating the development of clearer, more organized and verifiable specifications.