Towards requirements engineering for GenAI-enabled software: Bridging responsibility gaps through human oversight requirements
Abstract
Context: Responsibility gaps, long-recognized challenges in socio-technical systems where accountability becomes diffuse or ambiguous, have become increasingly pronounced in GenAI-enabled software. The generative and adaptive nature of such systems complicates how human oversight and responsibility are specified, delegated, and traced. Existing requirements engineering approaches remain limited in addressing these phenomena, revealing conceptual, methodological, and artifact-level research gaps in analyzing responsibility gaps from a human oversight requirements perspective. Objective: This study aims to analyze these research gaps in the context of GenAI-enabled software systems, establishing a coherent perspective for systematic analysis of responsibility gaps from a human oversight requirements standpoint. Methods: The proposed design methodology is structured across three analytical layers — conceptualization, methodological, and artifact. A user study compared the proposed methodology with a baseline goal-oriented requirements engineering approach. Results: Participant-perception results indicated generally favorable views, while expert ratings of the generated requirement artifacts showed higher mean scores for six of the seven artifact-quality items, with the strongest descriptive improvements observed for consistency and representational traceability. Conclusion: The study proposes a methodology for understanding, identifying, and documenting responsibility gaps in GenAI-enabled software from a human oversight requirements perspective.