This article puts forward solutions to some of the ethical and legal dilemmas posed in the current discussion on how to program crash algorithms in autonomous or selfdriving cars. The first part of the paper defines the scope of the problem in the criminal legal field, and the next section gives a critical analysis of the proposal to always prioritise
the interest of the occupant of the vehicle in situations with conflict of interests. The principle of minimizing social damage as a model for configuring self-driving cars is examined in the third section. Despite its apparent plausibility, within th framework of a liberal legal system that recognises humans as free agents who have rights and responsibilities, maximizing the function of social utility does not justify harmful interference into a person’s legal sphere. Therefore, in the fourth part, the author argues the need to program the crash algorithms of autonomous cars based on a deontological understanding of the
system of justifications in cri inal law. The solution to the dilemma lies in a prior analysis of the legal positions of all agents involved in the conflict, from a perspective of the principles of autonomy and solidarity as the core of the system of ustifications.

EB00000004215K

Available