Alertness to AI Algorithmic Biases in the Healthcare System

Abstract

Artificial intelligence (AI) is assisting practitioners in performing certain tedious tasks with computer algorithms, particularly those involving large volumes of data, bringing recent progress in medicine. Technical and social biases can undermine confidence in these algorithms, so the algorithms still require human validation. Especially when the approach by which AI systems are made can cause ethical concern. This work aims to review how the concepts of biases are integrated in different countries. Particularly, the approach of the ethical dimension and regulation that France, the United States (US), and China have when developing AI should be studied to understand better how it can lead to biases and ethical interrogations. The purpose is to assess the basis on which algorithms are created in the present healthcare system. Results show (i) a necessity to enrich AI systems, increase their explainability, and recommendations. (ii) Countries’ approaches to AI’s opportunities align with their beliefs.

Ähnliche Arbeiten