Document ID: CGR-ETH21 | Last Updated: Apr. 28, 2021
The topic of ethics and responsibility comes up frequently in the context of Artificial Intelligence, and for good reason. The concept of intelligent machines is scary to many people. Others are just as concerned about the increasing power that humans have over data and the potential for malicious use of powerful technology.
To address these concerns, organizations are producing frameworks to help classify the various ethics-related concerns for AI and ways in which people can maintain control over the use of potent AI technology. However, the area of ethics and technology is not one of clear definitions, concise terminology with well-defined boundaries, and areas of well-developed concepts. In this report, Cognilytica explores the range of ethical frameworks across the concepts of societal ethics, responsible use of AI, AI systemic transparency, governance of AI systems, and algorithmic interpretability and explainability. This report provides a basis for comparison across ethical concepts for the identified frameworks and identifies the strengths and potential weaknesses of different proposed ethical frameworks. Cognilytica also proposes a potential “normative” ethical framework covering all the concepts identified in this document, with concise boundaries between ethical concepts.