Abstract

Ethics for artificial intelligence (AI) is a topic of growing practical relevance. Many people seem to believe that AI could render jobs obsolete in the future. Others wonder who is in charge for the actions of AI systems they encounter. Providing and prioritizing ethical guidelines for AI is therefore an important measure for providing safeguards and increasing the acceptance of this technology. The aim of this research is to survey ethical guidelines for the handling of AI in the ICT industry and evaluate them with respect to their relevance. For this goal, first, an overview of AI ethics is derived from the literature, with a focus on classical Western ethical theories. From this, a candidate set of important ethical guidelines is developed. Then, qualitative interviews with experts are conducted for in-depth feedback and ranking of these guidelines. Furthermore, an online survey is performed in order to more representatively weight the ethical guidelines in terms of importance among a broader audience. Combining both studies, a prioritization matrix is created using the weights from the experts and the survey participants in order to synthesize their votes. Based on this, a ranked catalogue of ethical guidelines for AI is created, and novel avenues for research on AI ethics are presented.

Share

COinS