Abstract

A lot of attention as regards legal aspects of artificial intelligence (AI) is devoted to the issues of legitimacy of granting legal personality to it, responsibility for AI actions or the issue of rights to AI or the results of its operation. Less attention is drawn to the issues of AI accessibility. This is certainly due to the fact that there are more fears as regards AI possibilities and potential threats that may arise from the lack of control over AI systems. Those fears are by no means groundless. It seems, however, that in the debate on defining the legal framework for the functioning of AI, the discussion on ensuring control and responsibility when using AI systems should be accompanied by analyses concerning the rules of access to those systems. Those analyses should take into account the principles of equal access to AI systems, specifically by persons who are particularly vulnerable as regards social exclusion. The cumulated access to AI by a narrow group may be equally (and perhaps even more) dangerous than the lack of control over AI activities. This is due to the fact that good or bad intentions may be entertained by man, not AI, which by definition as an artefact is devoid of affections. The balanced policy of access to AI may contribute to optimal utilisation of its potential as well as benefits for man and society. The article is aimed at outlining the problem of access to AI from the viewpoint of the science of law, with special focus on the danger of exclusion of disabled persons and women. It also provides de lege ferenda solution proposals. Finally, it is the intention of the authors to stimulate a debate in the Polish juridical discourse on the problems of access to AI.