Legal regulation of AI and morality in the context of natural law and legal positivism

Arturas Grumulaitis
Vilnius University

Legal regulation of AI and morality in the context of natural law and legal positivism

Artūras Grumulaitis https://orcid.org/0000-0001-5356-5425 PhD Student, Faculty of Law Vilnius University Saulėtekio al. 9, I rūmai, LT-10222 Vilnius, Lithuania E-mail: arturas.grumulaitis@tf.stud.vu.lt

Scientific research problem. The dynamics of the investment in Artificial Intelligence (AI) and the rapidly growing use of its technology pose many new ethical and legal challenges. Even though the new technology stimulates creativity and contributes to the optimization of business processes, it can infringe copyright, create deep fakes, and mislead users by generating nonrealistic information. With the popularity of generative AI models rising at the end of 2022, the categories of “ethics” and “morality” are increasingly being discussed, raising the question of whether everything that AI provides us with is acceptable and tolerable. This implies the natural law paradigm’s idea that morality is the foundation that is needed for good legal (including AI) regulation.

Based on the essential natural law and legal positivism characteristics (*), legal doctrine, and the newest AI regulation initiatives in the EU, this paper seeks to clarify how the intrinsic morality of natural law influences the legal regulation on AI.

The aim and the subject of the work. The subject of the work is the relationship between morality and legal regulation of AI. The aim of the work is to assess the influence of moral values on the regulation of AI and to answer whether the proposed legal regulation (Artificial Intelligence Act) sufficiently reflects fundamental moral principles. The work purposefully refers to the provisions of two legal paradigms (natural law and legal positivism), characterized by different approaches to the role of morality in the legal system. Due to the limited scope of the work, this study does not examine legal realism or other social paradigms.

The objectives of the work. To achieve the goal, the following objectives are set: 1) to analyze and summarize the concept of artificial intelligence and the need for its regulation and proposed regulatory principles; 2) to reveal the most important ethical issues raised by artificial intelligence technology; 3) to analyze the role of morality in the legal regulation of artificial intelligence from the point of view of natural law and legal positivism; 4) to assess whether the proposed EU legal regulation (draft of the Artificial Intelligence Act) sufficiently reflects fundamental moral principles.

Research methods. Several methods will be used during the current scientific research: the linguistic method, revealing and defining concepts relevant to the work, explaining their meaning in the context of the considered legal paradigms; the comparative method, analyzing the peculiarities of different legal paradigms formed in the Western legal tradition; the systematic method, analyzing scientific literature and legal acts, to reveal the content and meaning of legal norms.

Conclusions. Analyzing whether the whether the proposed legal regulation (Artificial Intelligence Act) sufficiently reflects fundamental moral principles, the following statements can be concluded:

  1. The rapid development of AI undoubtedly brings many advantages, but this technology also raises serious ethical challenges: impermissible discrimination of individuals; privacy and personal data protection violations; unauthorized (illegal) collection and use of data for commercial purposes; manipulation of opinions and choices; unauthorized use of copyrighted works in the algorithm training databases etc. These challenges arises due to the complexity, opacity and autonomy of AI systems. Therefore, when considering the new regulation, the greatest attention should be paid to the more detailed analysis and control of these characteristics, thus aiming to protect basic human rights;

  2. Most of the challenges posed by AI are related to the human values, so the moral aspect of law assessment is significant and could serve to create a more humane, fairer law. The biggest problem in applying the principles of natural law to today’s legal regulation is that the morality is an individual category of a person, so it is practically impossible to establish a coherent and consistent system of moral norms that could be transferred to the positive law, so the search for the best way of doing this, is a continuous process.

  3. In assessing today’s legal regulatory challenges in the field of artificial intelligence, we would probably find it difficult to apply a “pure legal theory” with a strictly hierarchical structure of norms, which is inflexible and would hardly fulfill its functions as technology changes. A strict separation of law from the morality would likely complicate the legal investigation of technology’s impact on human rights. The application of “soft positivism” has a better perspective in the context of technology regulation, but also requires a more detailed study of the relationship between law and morality, the search for modern “conventional morality” and efforts to integrate it into the modern legal system;

  4. The draft of the AI Act seeks to incorporate moral principles in the text. The development of ethical AI is based on the respect of human rights and fundamental freedoms enshrined in the Treaty of the European Union and the EU Charter of Fundamental Rights. However, the regulation could pay more attention to ensuring the legal requirements for the training data of algorithms (the transparency obligation provided in the draft now is too abstract), the status of output data and the disclosure of such data in relation to GDPR requirements. It is doubtful whether the regulation is late, because the technology is evolving, and the AI regulation will not come into force until 2024-2025. However, by enshrining basic moral principles in the regulation and strengthening the protection of basic human rights, we should be able to overcome these challenges more easily.

Keywords: artificial intelligence, legal positivism, natural law, ethics, moral, legal paradigm

(*) the work analyzes famous authors of natural law and legal positivism (St. Augustine, T. Aquinus, J. Finnis, L. L. Fuller, H. L. A. Hart, I. Kant, H. Kelsen)