The ethics of Artificial Intelligence

Abstract

The recent introduction of AI tools in the justice sector poses several ethical implications as risks for judges’ independence and for procedural transparency, and discrimination biases. By developing ethical frameworks governing AI application, private and public agents have been increasingly dealing with risks pertaining to the use of AI. By inventorying and analyzing a set of ethical documents through content analysis, this study highlights the ethical implications involved in the application of AI. Moreover, by investigating the CEPEJ Charter (European Commission for the Effectiveness of Justice of the Council of Europe), the unique ethical document focusing on AI in justice, we were able to clarify potential differences between justice and other contexts of AI application with respect to risks prospected and the protection of ethical principles. The analysis confirms that the discipline of AI is a complex subject that involves very different aspects and therefore needs a broad focus on all contexts of application.

An Inventory of AI ethics: Tracing 100 documents

“…Phase 1 -Data collection: Using human manual coding, we searched the web for the following keywords: "AI ethical frameworks," AI guiding principles, African AI ethical frameworks, European AI ethical frameworks, and International AI ethical frameworks. In addition, we reviewed the ethical documents' reference lists and inventories of other scholarly articles such as [9] and [8]. In this phase we identified 100 documents drafted on AI ethics and principles.…”

“…As a result, multiple normative frameworks, standards, and sets of ethical principles guiding the use of AI have emerged. These frameworks aim to develop norms or criteria that, if followed, should limit the threat that artificial intelligence poses to fundamental rights and ideals [7] [8].…”

“…In 2015, the first AI ethics documents started to emerge to address the negative implications of AI and guide the development of AI systems toward the greatest public good [5] [9] [6] [8]. Several frameworks for AI ethics, both general and sector specific, have arisen in the years thereafter to address the harm that can be caused to people and society by the misuse, abuse, bad design, or unforeseen negative effects of AI systems [3] [10].…”

“…Previous research attempted to investigate these ethical frameworks from specific perspectives. For instance, [8] examined a list of ethical frameworks from 2016 to 2019 in relation to the consequences of AI in justice systems. [8] analyzed documents of AI ethics with the aim of determining which ethical principles and risk factors the texts primarily converge on.…”

“…For instance, [8] examined a list of ethical frameworks from 2016 to 2019 in relation to the consequences of AI in justice systems. [8] analyzed documents of AI ethics with the aim of determining which ethical principles and risk factors the texts primarily converge on. The study findings demonstrate the potential distinctions between justice and other contexts of AI deployment in terms of anticipated risks and ethical values protected.…”