Free beta access. 50 free AI tokens each week. Optional top-ups available. See what is included
Opening Word document…

Tax nomenclature solves.

Popovič ’22 [Adrián; 2022; researcher at the Department of Financial Law, Tax Law and Economy, Faculty of Law, Pavol Jozef Šafárik University; Financial Law Review, “Taxation of Robots and AI – Problem of Definition,” vol. 25 no. 1]

We have outlined above the different ways in which AI or Robots can be defined for the purposes of legal regulation. At the same time, we stated that these approaches may differ in the definition of AI and Robots for tax purposes.

In this section, we would like to mention one of the possible approaches to the definition of the term "Robot" for tax purposes. For this purpose, we propose the use of the robot tax nomenclature.

The classification in the tax nomenclature would be made on the basis of two attributes. The first attribute is based on the definition by use. Thus, in the tax nomenclature of robots, robots would be classified separately according to the purpose for which the robot can be used, for example: production robots, mining robots, agricultural robots, medical robots, etc. If a particular robot were intended for several different uses, it would be assigned several nomenclature numbers under which such a robot could belong. In the case of a universal multipurpose robot (i.e., a robot that can perform a different activity depending on its program, or in principle any activity), it would have a separate nomenclature number and its taxation would depend on its actual use (which would be demonstrated in a credible manner when declaring the performed activity of such a robot, which it would possibly record and process itself).

The second element for inclusion in the tax nomenclature of robots would be the degree of autonomy of the robot concerned (i.e., in this respect it would be an element based on the definition according to capabilities). In this way, the tax nomenclature of robots would make it possible to classify robots according to the need for cooperation with the human element. Three basic categories of robots come into consideration in this regard, for example: a robot with a human operator (necessary constant supervision and guidance of the robot's activity); a robot in which the human operator must make an initial setting when changing activities (for example, enter a design that the robot then implements - typical example may be a CNC machine); robot without human operator.

We believe that the tax nomenclature offers several benefits. A clear classification of a certain type of robot in the category of the tax nomenclature of robots under a certain nomenclature number would make it possible, without further doubt, to define such an object of taxation in the causal world. Considering the parameter of cooperation with the human element would make it possible to subject to progressive taxation those robots which are characterized by a higher element of autonomy compared to human operators. This would lead to an increase in tax revenues, especially where there is a reduction of tax revenues from the taxation of human labor.

For a better imagination, within the selection of a specific variant of the possibility of such taxation, it would be possible to create a graph consisting of two vectors, while the horizontal vector (X axis) would represent a set of tax objects (taxation according to the purpose for which the robot can be deployed) and the vertical vector (Y axis) would express a set expressing at each point of the X-axis a specific classification according to the extent and scope of taxation of an individual taxation object (taxation based on abilities graded, for example, according to the need for cooperation with the human element). By substituting specific data on individual vectors, it would be possible to monitor the development of the tax burden of individual tax objects classified by purpose with regard to the ability of the tax object and the need for cooperation with the human element (Graph 1).

<<GRAPH OMITTED>>

The basic question when classifying robots in the tax nomenclature is who should be the authority that decides on this classification. We would not recommend leaving this decision to the discretion of the tax administrator – as this would conflict the AI/R tax with the principle of legality and legal certainty for the addressees of tax legal norms.

A suitable alternative seems to be the transfer of the obligation to classify robots into individual classes to robot manufacturers, as it is these persons who have (and should) know the parameters of the robots they manufacture (which is the most rational assumption). Persons producing robots in a certain volume, resp. with a certain turnover, they would be required to register as robot taxpayers. In addition to registration obligations, with this system would be associated evidential obligations (including the assignment of a nomenclature number to a certain robot) or the obligation to collect the tax applied on the basis of tax - the price of the robot (in case the legislator decides to indirectly tax robots with excise duty). In the event that this classification proves to be incorrect (inspection by the tax administrator of a customer using a robot for a purpose other than that specified by the manufacturer), these persons could also be the guarantors of such a robot tax.

In connection with the above mentioned, the question arises whether the proposed transfer of the obligation to classify to manufacturers is not just a circular definition - that is, how do we know which entity manufactures robots, when it is this entity that should classify certain physical units as robots? In fact, this is not a definition of a circle. The condition setting the registration obligation for a certain entity would be based on a different definition than the subsequent inclusion of robots in the tax nomenclature. In this regard, a condition based on capability definition seems appropriate. With reference to what we have mentioned above, a robot manufacturer could be, for example, a person producing a unit acquires autonomy through sensors and/or by exchanging data with its environment (inter-connectivity); trades and analyses data; is self-learning (optional criterion); has a physical support; adapts its behaviours and actions to its environment. A definition of a more general nature seems to be applicable in this case, as the radius of these persons could be further narrowed by additional criteria (such as turnover and / or number of produced robots). Subsequently, the robot manufacturers thus registered would carry out the actual classification of the robots produced for the purposes of the robot tax nomenclature.

In the above-mentioned way, robots would become object of taxation already during their production, resp. when the robot is "released for free circulation" for the first time3.

Sign in to recut, update, or save this card.