Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Implementing Responsible AI: Tensions and Trade-Offs Between Ethics Aspects

Conrad Sanderson, David Douglas, Qinghua Lu

TL;DR

In this work, a catalogue of 10 notable tensions, trade-offs and other interactions between the underlying aspects of ethics principles are compiled and discussed, focusing on two-sided interactions.

Abstract

Many sets of ethics principles for responsible AI have been proposed to allay concerns about misuse and abuse of AI/ML systems. The underlying aspects of such sets of principles include privacy, accuracy, fairness, robustness, explainability, and transparency. However, there are potential tensions between these aspects that pose difficulties for AI/ML developers seeking to follow these principles. For example, increasing the accuracy of an AI/ML system may reduce its explainability. As part of the ongoing effort to operationalise the principles into practice, in this work we compile and discuss a catalogue of 10 notable tensions, trade-offs and other interactions between the underlying aspects. We primarily focus on two-sided interactions, drawing on support spread across a diverse literature. This catalogue can be helpful in raising awareness of the possible interactions between aspects of ethics principles, as well as facilitating well-supported judgements by the designers and developers of AI/ML systems.

Implementing Responsible AI: Tensions and Trade-Offs Between Ethics Aspects

TL;DR

In this work, a catalogue of 10 notable tensions, trade-offs and other interactions between the underlying aspects of ethics principles are compiled and discussed, focusing on two-sided interactions.

Abstract

Many sets of ethics principles for responsible AI have been proposed to allay concerns about misuse and abuse of AI/ML systems. The underlying aspects of such sets of principles include privacy, accuracy, fairness, robustness, explainability, and transparency. However, there are potential tensions between these aspects that pose difficulties for AI/ML developers seeking to follow these principles. For example, increasing the accuracy of an AI/ML system may reduce its explainability. As part of the ongoing effort to operationalise the principles into practice, in this work we compile and discuss a catalogue of 10 notable tensions, trade-offs and other interactions between the underlying aspects. We primarily focus on two-sided interactions, drawing on support spread across a diverse literature. This catalogue can be helpful in raising awareness of the possible interactions between aspects of ethics principles, as well as facilitating well-supported judgements by the designers and developers of AI/ML systems.
Paper Structure (20 sections, 1 figure)

This paper contains 20 sections, 1 figure.

Table of Contents

  1. Introduction
  2. Underlying Aspects
  3. Accuracy
  4. Robustness
  5. Fairness
  6. Privacy
  7. Explainability
  8. Transparency
  9. Interactions between Aspects
  10. Accuracy vs. Robustness
  11. Accuracy vs. Fairness
  12. Accuracy vs. Privacy
  13. Accuracy vs. Explainability
  14. Fairness vs. Robustness
  15. Fairness vs. Privacy
  16. ...and 5 more sections

Figures (1)

  • Figure 1: Summarised form of the high-level AI ethics principles proposed by the Australian Government AusAIprinciples_2019.