SureBuilt Series
A Human-Centric Governance Framework for Responsible Human-AI Collaboration in The Built Environment Systems: Integrating Human and Machine Epistemologies Within
- Department of Architecture, Caleb University, Lagos, Nigeria; Works Department, Yaba College of Technology, Yaba, Lagos, Nigeria; Phraysal Nigeria Limited, Sagamu, Ogun State, Nigeria
Abstract
Artificial intelligence (AI) systems increasingly mediate built-environment decisions. Yet, governance frameworks still say more about system risk than about why a particular AI-mediated claim should gain authority in a consequential decision. This conceptual article develops epistemic governance as the institutional ordering of how material decision claims are constituted, credited, admitted, challenged, closed and revised. A targeted theory-synthesis review, completed on 11 August 2026, integrates responsible-AI governance, human oversight, actor-network theory (ANT), epistemic injustice, Industry 5.0 and built-environment decision research. ANT explains how authority is assembled through translation, inscription, composition, delegation and black-boxing; epistemic injustice theory evaluates credibility, standing, exclusion and contestability. Their synthesis yields the Epistemic Authority Assurance Protocol (EAAP), comprising five claim-level checks: claim constitution, representational standing, inferential warrant, contestable uptake, and accountable closure and learning. Seven reversible dispositions specify whether a claim may be authorised, conditioned, held, revised, rejected, escalated or reopened. A stylised municipal flood-prioritisation illustration shows how sparse data and monetised loss can make a technically plausible ranking epistemically incomplete. The contribution is a portable, auditable protocol that complements rather than replaces established AI risk management and legal frameworks. AI remains a producer and mediator of propositions; accountable natural and legal persons retain decision responsibility. Leveraging Generative AI for Carbon-Neutral Building Envelopes in the Digital Age Vincent Chubo¹*, Deliverance Chiripanyanga¹, Panashe Mashonganyika¹ Quantity Surveying, National University of Science and Technology, Bulawayo, Zimbabwe Chipondavincent84@gmail.com The building envelope plays a central role in heat transfer, natural lighting, natural ventilation, and whole-life carbon emissions, yet many BIM workflows still rely on sequential, single-variable iteration that limits exploration of competing design objectives. This paper compar
Retrieved from https://surebuiltseries.com/papers/a-human-centric-governance-framework-for-responsible-human-ai-collaboration-in-the-built-environment-systems-integrating-human-and-machine-epistemologies-within/