An AI policy ontology for the civil security domain

Abstract Artificial Intelligence (AI) has the potential to unlock new opportunities for data-driven and evidence-based policy making. However, there is still a lack of knowledge and understanding regarding the potential use and related impact of AI-based technologies in civil security, which often lead to partially-informed perspectives and distrust. Ethical and secure-by-design algorithms as well as a broader engagement of civil society on the values to be embedded in AI are necessary to build trust in this technology, especially in the Law Enforcement (LE) domain. This paper focuses on the provision of concrete, value-adding, and holistic policy recommendations that effectively respond to the new technology-enabled environment of fast paced-change, taking into consideration the multiple voices of the society. Contributing to the ethical and responsible AI, this paper introduces an AI policy ontology serving as a conceptual and technical framework to advance the development of trustworthy and socially acceptable AI within the civil security domain. It supports the analysis of key AI policy concepts and players, offering a high-level, unified perspective and a mapping tool for developing AI policies. This analysis is subsequently applied in the context of the European Union (EU) civil security ecosystem in order to develop and map a set of concrete policy recommendations aiming to promote ethics and civil rights while fostering trust in AI systems for Law Enforcement.