Human factors in the standardization of AI governance: Improving the design of risk management standards for ethical AI

Kilhoffer, Zachary

Permalink

https://hdl.handle.net/2142/129214 Copy

Description

Title

Human factors in the standardization of AI governance: Improving the design of risk management standards for ethical AI

Author(s)

Kilhoffer, Zachary

Issue Date

2025-04-24

Director of Research (if dissertation) or Advisor (if thesis)

Sanfilippo, Madelyn

Doctoral Committee Chair(s)

Wang, Yang

Committee Member(s)

• Bashir, Masooda
• Ma, Jiaqi

Department of Study

Information Sciences

Discipline

Information Sciences

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• AI Governance
• Trustworthy AI
• Ethical AI
• AI Standards

Language

eng

Abstract

The field of AI governance is rapidly standardizing, with notable examples such as the NIST AI Risk Management Framework (AI RMF) and ISO/IEC 42001 illustrating a standardized risk management approach. However, a significant disconnect persists between academic research and practitioner needs, with the latter relying on industry standards for real-world implementation of ethical principles. These emerging standards, while promising, remain untested and face numerous challenges in effectively governing the complex and dynamic field of AI. This dissertation examines the standardization of AI governance, adopting a constrained optimization approach to improve AI standards. The research addresses three key questions: (1) What are the system-level requirements for standardization as a governance framework for human-centered AI systems? (2) What are the unit-level requirements for standards as a governance instrument for human-centered AI systems? (3) How can AI Risk Standards be designed to achieve better outcomes in the development of human-centered AI systems? Through three empirical contributions, this work aims to enhance emerging AI standards by focusing on the standards themselves, the AI practitioners who will implement them, and the organizational contexts in which they operate. The research provides practical guidance for AI practitioners tasked with implementing AI standards and offers recommendations for standard developing organizations. By addressing the challenges of AI governance through standardization, this dissertation contributes to the development of more effective, human-centered AI systems. It bridges the gap between theoretical principles and practical implementation, offering insights that can shape the future of AI governance and ensure the responsible development of AI technologies.

Graduation Semester

2025-05

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/129214

Copyright and License Information

Copyright 2025 Zachary Kilhoffer

Owning Collections