Problems of Legal Regulation of Algorithmic Administrative Acts in Georgia

Abstract

The rapid development of information technologies raises the need for the introduction of algorithmic administrative acts. Algorithmic decisions, as an administrative act, affect fundamental human rights. An algorithmic administrative act is a decision made by an administrative body, in the process of which the administrative body relies on a software algorithm.

Although the final decision can be approved by a person or issued automatically, in legal reality it is still considered an act issued by an administrative body, since the legal consequences of the decision create, change or terminate the rights and obligations of a person. Although digital governance in Georgia is developing gradually, we do not have a unified legal framework and what we have fragmentarily considers the specifics of automated decisions, which requires both changes in legislation and the creation of additional guarantees of transparency, ethics and protection of human rights.

It is worth noting that the current administrative procedural law does not take into account the specifics of a decision made by an algorithm, in particular, it is not determined whether an automated decision constitutes an administrative act; the limits of responsibility of the algorithm developer, the administrative body and the final decision-making entity are unclear; There is no standard for the level of transparency and justification of algorithms; citizens have limited opportunities to substantiate and appeal automated decisions, the procedure for which is unknown to them.

In the research process, it is important to find out how the legal regulation of algorithmic administrative acts should be organized in Georgia in the conditions when state structures are gradually switching to automated decisions. Examples of transport optimization in Tbilisi, social assistance allocation, risk-based customs control and other areas demonstrate that automated decisions are already being made by administrative bodies. Against the background of technological progress, decisions that are analyzed not by a person, but by an algorithm are becoming increasingly common, which raises new challenges both in terms of protecting fundamental human rights and state accountability.

Accordingly, the Georgian legal system faces the challenge of how to adopt technological capabilities without violating fundamental human rights and the principles of fair administrative proceedings.

The study aims to develop a unified view of whether an algorithmic decision constitutes an administrative act in the classical sense; how the right to justification works when the justification is built into the “code”; to what extent is it possible for a person to appeal a decision when the decision mechanism itself is unclear to them; and how does the state protect the security of personal data when decisions are based on automatic processing or high-risk AI models.

The essence of the problem lies in the fact that Georgia does not yet have a unified, clear, specialized legal regulation that would include the identification, regulation, transparency, justification, human involvement and accountability of algorithmic administrative acts. The Administrative Procedure Code of Georgia does not contain special regulations on algorithmic decisions. The use of artificial intelligence in administrative activities raises questions about legality, transparency, accountability and protection of human rights. Also, in the context of European integration, Georgia is obliged to bring national legal acts into line with the standards of the EU AI Act.