Artificial Intelligence (AI) is one of the most strategic technologies of the 21st century. Due to the rapid emergence
of new AI based systems, its growing advancement, and growing computer power, availability of data and
progress in algorithms, many countries and the European Union (EU) as a whole, are now facing the challenge
to address issues inextricably linked with this new technology. AI technology is both transformative and
disruptive. The emergence of AI-based systems poses numerous legal and ethical questions. There are many
kinds of risks as well as opportunities which are connected to this technology. Therefore, legislators should
impose a legal framework that will handle the risks and ensure that AI systems will be beneficial to society.
Both the EU and United States of America (US) are in the midst of creating their regulations on AI-based systems.
The comparison of the EU and US efforts towards AI regulation in this thesis highlights the similarities and
differences in their respective approaches, which in turn may help people understand what issues both sides
need to address and how to improve their efforts. The goal of this thesis is to introduce and assess potential legal
challenges European and American legislators face and recommend necessary solutions. This thesis focuses on the most urgent legal issues that are within the scope of the powers of the EU and the US
federal government which are currently being heavily discussed by both parties. The first chapter of the thesis
focuses on the beginnings of the field of AI, its modern history, and definition. The second chapter presents
current strategies that are being developed by the EU and US. The thesis describes the current legislative efforts
of the European Parliament, the European Commission, European organizations, the US White House and its
Congress, and non-governmental organizations. The third chapter deliberates extra-contractual liability for AI based
system operations, especially it identifies liable persons for damage caused by AI systems, describes
possible standards of liability, and presents issues connected to the burden of proof. In the fourth chapter, the
thesis addresses potential legal issues involving product liability for AI based systems and necessary
recommendations are given. The fifth chapter addresses the proposition of equipping AI-based systems with
legal personhood. The sixth chapter describes various propositions for ethical and legal guidelines of AI R&D.
