SAFEARC: A Seven-Pillar Operational Architecture for Socio-Technical AI Governance
Year: 2025
Type: article
Abstract: Despite broad convergence among global AI governance frameworks, organizations continue to struggle with implementation. The barrier is not conceptual but operational: existing guidance defines principles without translating them into workable socio-technical processes. This paper introduces SAFEARC, a seven-pillar architecture designed to operationalize regulatory expectations across organizational contexts. SAFEARC integrates requirements from the EU AI Act, OECD AI Principles, NIST AI Risk Management Framework, and ISO/IEC 42001 into a unified workflow structured around Scan, Align, Filter, Evaluate, Assign, Renew, and Contain.Using clause-level crosswalk analysis, convergence mapping, and synthesis across 68 regulatory and standards documents, the study identifies six governance domains consistently demanded by major frameworks and four structural barriers that impede adoption in practice. These findings inform the development of an architecture aimed at bridging the gap between policy intent and organizational capability.The paper demonstrates that SAFEARC enhances auditability, improves internal coordination, and supports socio-technical readiness for emerging AI regulation. The approach is applicable to both large enterprises and resource-constrained organizations, offering a practical pathway toward trustworthy AI implementation.
Author: Carvalho, Waydell D.
Language: English
Subfield: Safety Research
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7116685346
FWCI:
Cites:
Cited by:
Related to: