What do we know about implementing ethical principles in the context of Artificial Intelligence? A Systematic Literature Review

Context: In recent times, there has been a notable increase in academic discourse surrounding the ethical implications of artificial intelligence (AI). This has been driven by the exponential growth in both the usage and development of AI, which is now an integral component of modern technology. However, even with the ongoing research in this area, the issue of AI ethics remains a source of contention, as evidenced by the ongoing debates among researchers. Although ethical issues such as fairness and privacy have been the subject of study, their application in the context of AI development remains uncertain. Goal: The aim of this paper is to identify the current state of ethical AI research in order to gain insight into the challenges encountered in integrating ethics into AI development and the proposed strategies to address these challenges. Methodology: To address this, we conducted an update to the Systematic Literature Review (SLR) published by Cerqueira in 2021. We evaluated research published until April 2023 across three databases to answer three research questions. Results: The collection yielded 1712 studies, of which 38 were determined to be appropriate for analysis. The majority of studies (63%) propose a practical solution to facilitate the implementation of AI ethics. Furthermore, we were able to identify a list containing 26 major ethical principles that have been discussed throughout the literature. Conclusion: The review revealed a considerable diversity in the methods, processes and suggestions proposed, with regard to the type of approach, life cycle phase and objectives. This indicate that, despite the current limitations of AI ethics, the field is developing and addressing a greater number of solutions to mitigate the risks and ethical problems present in AI systems. Keywords: Ethics, AI ethics, Artificial Intelligence, AI ethics principles, Ethical requirements