This thesis investigates the possibilities and boundaries of generative Artificial Intelligence (AI) to conceptual reinterpret Iranian neighbourhoods (Mahallehs) in the regime of vertical housing. Contemporary Iranian cities are characterized by urban densification and widespread application of high-rise housing, which often produces placeless environments that do not respect cultural identity, social bonds, and climatic appropriateness. Conversely, the old Mahalleh embodied socio-spatial notions such as hierarchy, privacy, communal interaction, and responsiveness to the environment. This research therefore asked: to what degree, and with what constraints, can new urban settings re-elaborate these notions using generative AI? The study applied a design-based methodology combining cultural grounding, iterative prompt engineering with Midjourney, and a systematic evaluation framework of eight categories that emerged from principles in everyday life.