The Paradigm Shift in Architecture: A Systematic Review of Artificial Intelligence Integration (2022–2026)
The Paradigm Shift in Architecture: A Systematic Review of Artificial Intelligence Integration (2022–2026)
Saman Mohri1
1) Adjunct Lecturer, Department of Architecture and Urban Planning, Technical and Vocational University (TVU), Shiraz, Iran,
IOI Code
:
AACZ-ABHCAK
Publication :
The 5th International Conference on Seismic Retrofitting(5icst.ir)
Abstract :
The present study constitutes a systematic review of the paradigm metamorphosis that has occurred between 2022 and 2026 as a result of integrating artificial intelligence with the architectural design process. The foundational objective of this review is to formulate a coherent conceptual framework for understanding the transition from rule-based parametricism to autonomous agentic systems, and to elucidate the human-centric nature of this transformation. The research methodology is grounded in the standard PRISMA protocol, and out of one thousand two hundred and eighty-four records retrieved from the Scopus and Web of Science databases, sixty-one landmark studies with the highest methodological quality were selected.
The findings indicate that this metamorphosis is organized into five distinct yet interconnected pillars: generative design and diffusion models, spatial analytics and feasibility automation, multimodal large language models and agentic workflows, environmental and performance-driven optimization, and finally, environmental psychology and perceptual measures. By exploring the mathematical mechanism of diffusion models, the multi-agent syntactic parsing of Building Information Modeling (BIM), and particularly the quantification of eye-tracking metrics such as fixation duration and saccade frequency, this review demonstrates that a profound technological convergence is emerging, wherein probabilistic visual artificial intelligence synthesizes with deterministic structural and environmental verification engines and human perceptual measures. The most significant theoretical contribution is the elucidation of the cognitive-spatial optimization loop, in which bio-perceptual data is transformed into a feedback signal to guide artificial intelligence generation. The results indicate that the novel paradigm does not lead to the replacement of the architect, but rather to the augmentation of their agency and the realization of a responsive, human-centric, and resilient built environment, contingent upon the meticulous management of the data synchronization challenge and ethical considerations.
Keywords : generative artificial intelligence (generative ai) large language models (llms) architectural design performance driven optimization systematic review