Ethical and Data Protection Gaps in the Development and Review of Artificial Intelligence Solutions in Healthcare

Abstract

The increase in development of solutions for healthcare using Artificial Intelligence (AI) has led to better health care outcomes in the recent past. AI solutions have however brought in challenges in how to safely use them on patients while at the same time ensuring that all the ethical principles are upheld. This study sought determine: (a) the level of understating and application of ethical principles by AI researchers and practitioners during the development of machine learning (ML) systems in health care, (b) the level of understanding of ML by reviewers of select Institutional Ethics review boards. Two focus group discussion (FGD) made up of six members each were conducted. one with AI/ML practitioners and researcher and the other with scientific and research ethics committee members. Findings from the FGD indicate that AI/ML practitioners are not adequately aware of the ethical challenges in AI/ML. Scientific and research ethic committee members lack knowledge on AI/ML and are not able to effectively review AI/ML proposal and projects. Most members from Both AI/ML practitioners/researchers and ethics committee members are not well conversant with the data protection act of 2019. Recommendations include: (a) training of AI/ML practitioners/researcher on ethics in AI (b) train ethics committee members on the fundamentals of AI/ML (c) train both groups on data protection act of 2019.