Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Navigating Fairness: Practitioners' Understanding, Challenges, and Strategies in AI/ML Development

Aastha Pant, Rashina Hoda, Chakkrit Tantithamthavorn, Burak Turhan

TL;DR

This paper empirically investigates AI practitioners' understanding of fairness in AI/ML, their development challenges, perceived consequences of unfair systems, and practical strategies to promote fairness. It uses 22 semi-structured interviews and Socio-Technical Grounded Theory (STGT) to build a framework linking practitioners' fairness concepts to challenges, outcomes, and mitigation approaches. Key findings show that dataset access and data bias are central bottlenecks, while practitioners converge on bias mitigation as a core fairness strategy, regardless of their definitional stance. The work offers actionable recommendations for practitioners and organizations to balance fairness with product delivery and suggests future research directions to align policy definitions with real-world practice and tools.

Abstract

The rise in the use of AI/ML applications across industries has sparked more discussions about the fairness of AI/ML in recent times. While prior research on the fairness of AI/ML exists, there is a lack of empirical studies focused on understanding the perspectives and experiences of AI practitioners in developing a fair AI/ML system. Understanding AI practitioners' perspectives and experiences on the fairness of AI/ML systems are important because they are directly involved in its development and deployment and their insights can offer valuable real-world perspectives on the challenges associated with ensuring fairness in AI/ML systems. We conducted semi-structured interviews with 22 AI practitioners to investigate their understanding of what a 'fair AI/ML' is, the challenges they face in developing a fair AI/ML system, the consequences of developing an unfair AI/ML system, and the strategies they employ to ensure AI/ML system fairness. We developed a framework showcasing the relationship between AI practitioners' understanding of 'fair AI/ML' system and (i) their challenges in its development, (ii) the consequences of developing an unfair AI/ML system, and (iii) strategies used to ensure AI/ML system fairness. By exploring AI practitioners' perspectives and experiences, this study provides actionable insights to enhance AI/ML fairness, which may promote fairer systems, reduce bias, and foster public trust in AI technologies. Additionally, we also identify areas for further investigation and offer recommendations to aid AI practitioners and AI companies in navigating fairness.

Navigating Fairness: Practitioners' Understanding, Challenges, and Strategies in AI/ML Development

TL;DR

This paper empirically investigates AI practitioners' understanding of fairness in AI/ML, their development challenges, perceived consequences of unfair systems, and practical strategies to promote fairness. It uses 22 semi-structured interviews and Socio-Technical Grounded Theory (STGT) to build a framework linking practitioners' fairness concepts to challenges, outcomes, and mitigation approaches. Key findings show that dataset access and data bias are central bottlenecks, while practitioners converge on bias mitigation as a core fairness strategy, regardless of their definitional stance. The work offers actionable recommendations for practitioners and organizations to balance fairness with product delivery and suggests future research directions to align policy definitions with real-world practice and tools.

Abstract

The rise in the use of AI/ML applications across industries has sparked more discussions about the fairness of AI/ML in recent times. While prior research on the fairness of AI/ML exists, there is a lack of empirical studies focused on understanding the perspectives and experiences of AI practitioners in developing a fair AI/ML system. Understanding AI practitioners' perspectives and experiences on the fairness of AI/ML systems are important because they are directly involved in its development and deployment and their insights can offer valuable real-world perspectives on the challenges associated with ensuring fairness in AI/ML systems. We conducted semi-structured interviews with 22 AI practitioners to investigate their understanding of what a 'fair AI/ML' is, the challenges they face in developing a fair AI/ML system, the consequences of developing an unfair AI/ML system, and the strategies they employ to ensure AI/ML system fairness. We developed a framework showcasing the relationship between AI practitioners' understanding of 'fair AI/ML' system and (i) their challenges in its development, (ii) the consequences of developing an unfair AI/ML system, and (iii) strategies used to ensure AI/ML system fairness. By exploring AI practitioners' perspectives and experiences, this study provides actionable insights to enhance AI/ML fairness, which may promote fairer systems, reduce bias, and foster public trust in AI technologies. Additionally, we also identify areas for further investigation and offer recommendations to aid AI practitioners and AI companies in navigating fairness.
Paper Structure (49 sections, 8 figures, 2 tables)

This paper contains 49 sections, 8 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Background and Motivation
  3. Definition and Approaches on 'AI/ML fairness'
  4. Review studies on AI/ML fairness
  5. Research Methodology
  6. Study Design
  7. Participant Information
  8. Understanding Participants' Perspectives and Experiences in Developing a Fair AI/ML System
  9. Pilot study
  10. Interview Sampling and Data Collection
  11. Data Analysis
  12. Findings
  13. Participants' Demographics
  14. RQ 1- What do AI practitioners understand by ‘fair AI/ML’?
  15. In terms of absence of bias
  16. ...and 34 more sections

Figures (8)

  • Figure 1: Overview of the research methodology of our study
  • Figure 2: Examples of STGT analysis hoda2021socio applied to qualitative data on the challenges in developing a fair AI/ML system.
  • Figure 3: An example of a memo on AI practitioners' understanding of 'fair AI/ML'
  • Figure 4: Overview of the participant's understanding of 'fair AI/ML'
  • Figure 5: Overview of the participants' challenges in developing a fair AI/ML system
  • ...and 3 more figures