Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Trustworthy and Responsible AI for Human-Centric Autonomous Decision-Making Systems

Farzaneh Dehghani, Mahsa Dibaji, Fahim Anzum, Lily Dey, Alican Basdemir, Sayeh Bayat, Jean-Christophe Boucher, Steve Drew, Sarah Elaine Eaton, Richard Frayne, Gouri Ginde, Ashley Harris, Yani Ioannou, Catherine Lebel, John Lysack, Leslie Salgado Arzuaga, Emma Stanley, Roberto Souza, Ronnie de Souza Santos, Lana Wells, Tyler Williamson, Matthias Wilms, Zaman Wahid, Mark Ungrin, Marina Gavrilova, Mariana Bento

TL;DR

The intricacies of AI biases, definitions, methods of detection and mitigation, and metrics for evaluating bias are reviewed, as well as guidelines to foster Responsible and Trustworthy AI models are discussed.

Abstract

Artificial Intelligence (AI) has paved the way for revolutionary decision-making processes, which if harnessed appropriately, can contribute to advancements in various sectors, from healthcare to economics. However, its black box nature presents significant ethical challenges related to bias and transparency. AI applications are hugely impacted by biases, presenting inconsistent and unreliable findings, leading to significant costs and consequences, highlighting and perpetuating inequalities and unequal access to resources. Hence, developing safe, reliable, ethical, and Trustworthy AI systems is essential. Our team of researchers working with Trustworthy and Responsible AI, part of the Transdisciplinary Scholarship Initiative within the University of Calgary, conducts research on Trustworthy and Responsible AI, including fairness, bias mitigation, reproducibility, generalization, interpretability, and authenticity. In this paper, we review and discuss the intricacies of AI biases, definitions, methods of detection and mitigation, and metrics for evaluating bias. We also discuss open challenges with regard to the trustworthiness and widespread application of AI across diverse domains of human-centric decision making, as well as guidelines to foster Responsible and Trustworthy AI models.

Trustworthy and Responsible AI for Human-Centric Autonomous Decision-Making Systems

TL;DR

The intricacies of AI biases, definitions, methods of detection and mitigation, and metrics for evaluating bias are reviewed, as well as guidelines to foster Responsible and Trustworthy AI models are discussed.

Abstract

Artificial Intelligence (AI) has paved the way for revolutionary decision-making processes, which if harnessed appropriately, can contribute to advancements in various sectors, from healthcare to economics. However, its black box nature presents significant ethical challenges related to bias and transparency. AI applications are hugely impacted by biases, presenting inconsistent and unreliable findings, leading to significant costs and consequences, highlighting and perpetuating inequalities and unequal access to resources. Hence, developing safe, reliable, ethical, and Trustworthy AI systems is essential. Our team of researchers working with Trustworthy and Responsible AI, part of the Transdisciplinary Scholarship Initiative within the University of Calgary, conducts research on Trustworthy and Responsible AI, including fairness, bias mitigation, reproducibility, generalization, interpretability, and authenticity. In this paper, we review and discuss the intricacies of AI biases, definitions, methods of detection and mitigation, and metrics for evaluating bias. We also discuss open challenges with regard to the trustworthiness and widespread application of AI across diverse domains of human-centric decision making, as well as guidelines to foster Responsible and Trustworthy AI models.
Paper Structure (11 sections, 2 figures, 1 table)

This paper contains 11 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Trustworthy and Responsible AI Definition
  3. Governance for Human-Centric Intelligence Systems
  4. Biases
  5. Strategies to Detect Biases
  6. Evaluating biases
  7. Bias Mitigation in Intelligent Systems
  8. Trustworthy and Responsible AI in Human-centric Applications
  9. Open Challenges
  10. Guidelines and Recommendations
  11. Conclusion and Final Remarks

Figures (2)

  • Figure 1: An overview of AI principles discussed in Section \ref{['sec:keywords']}. The diagram depicts the intricate relationship and intersections between fundamental concepts crucial for the development and deployment of AI systems. Emphasized are nine central tenets: Responsibility, Explainability, Trustworthiness, Fairness, Data Security, AI Governance, Reliability, Reproducibility, and Communication. Surrounding these core ideas are various sub-components, serving as the building blocks and considerations that further refine and give depth to each principle, emphasizing the comprehensive nature of ethically implementing AI.
  • Figure 2: The summary of sources of bias, bias detection, and bias mitigation techniques. a) the feedback loop of bias, wherein various biases that can intrude on AI systems during data collection and model development are depicted. b) a number of fairness evaluation metrics that can be applied to identify bias in AI systems. If the bias detected can be addressed, the model should be adjusted. For each specific problem one or more metrics can be applied in the Pre-processing, In-processing, and Post-processing stages. c) a design process to mitigate bias in the Pre-processing, In-processing, and Post-processing stages to improve Fairness in AI systems. d) post-authorization monitoring of data and AI system, through qualitative user studies or fairness evaluation metrics.