The Ethical Challenge of AI in Architecture: Formulating a Human-Centric Framework for Algorithmic Governance

Saman Mohri¹ Saeedeh Dehghani²

1) Ph.D. Candidate, Department of Architecture, Faculty of Art and Architecture, Shiraz Branch, Islamic Azad University, Shiraz, Iran.
2) Lecturer, Department of Architecture, Al-Zahra National Skills University, Bushehr Branch, Bushehr, Iran.

Publication : 4th.International Conference on Architecture, Civil Engineering, Urban Development, Environment and Horizons of Islamic Art in the Second Step Statement of the Revolution(4icacu.com)
Abstract :
The increasing convergence of Artificial Intelligence (AI) with architecture is transitioning from an instrumental upgrade to a new paradigm, wherein algorithmic systems emerge as active agents in cognitive, creative, and production processes. This fundamental transformation, alongside its potential benefits, introduces a series of complex ethical challenges. These perils are not merely technical errors but are the logical consequences of the unguided implementation of technologies that can inadvertently amplify historical biases, erode inhabitants privacy, and undermine human agency. The existing technical discourse lacks a systematic approach to confronting these issues. Aiming to bridge this gap, this paper first provides a meticulous analysis of the key ethical dimensions of this convergence, including bias and fairness, privacy and surveillance, agency and creativity, and transparency and accountability. Adopting a qualitative approach based on a critical content analysis of interdisciplinary literature, this research argues that effectively addressing these challenges necessitates a practical and conceptual guiding framework. In response to this necessity and to transition from mere critique to a constructive proposal, the paper designs a Human-Centric Framework for Guiding Algorithms. This framework is founded on four fundamental and interconnected principles: 1) Interpretable Transparency, which emphasizes making the system s logic understandable to build trust; 2) Distributed Accountability, which defines and clarifies responsibility among human agents in a project; 3) Contextual Fairness, which calls for the active evaluation and mitigation of discriminatory biases in data and algorithmic outputs; and 4) Interactive Agency, which insists on designing systems that offer suggestions to humans—rather than making decisions for them—preserving the possibility of final control and intervention. Ultimately, this research asserts that responsible intelligent architecture necessitates the evolution of the architect s role into that of an ethical steward who, by integrating critical insight with technical skill, places human values at the center of the algorithmic design process.
Keywords : artificial intelligence (ai) architectural ethics algorithmic governance human centric architecture digital ethics computational design