A Feasibility Study of Programmer Use of Necessity as a Defense for Self-Driving Cars Based on the Foundations of Iranian Criminal Law and Precedents Relative to Homicide-by-Necessity in Common Law

Abstract

Developments in transportation industry and computer science have caused the production of cars that do not need human drivers, and instead are driven by artificial intelligence. In this case, the programmer has to predict the situations in which a potential accident is expected previously and provide training to ensure the self-driving cars’ reactions are in compliance with the law. In possible scenarios, which could occur during driving, reacting in compliance with the law is a fundamental challenge that should be settled by lawyers so that the programmer set the self-driving car’s algorithms according to the lawyers’ proposal in a way that probable accidents do not lead to his/her criminal responsibility. Considering this fact, in the present article by using library resources and through a descriptive-analytic method, depiction of potential scenarios, including cases in which the self-driving car is set up to make a choice between hitting two distinct groups of people is dealt with. Finally, paying attention to foundations of domestic law and precedents relative to homicide-by-necessity in common law system, we are seeking to answer these challenges. As such, it is attempted to present the necessary theoretical perspective in order to make Iran’s legal system ready for the entry of self-driving cars and to reduce the risk for designers and investors’ activities in this field through eliminating the criminal responsibility caused by accident of self-driving cars with human based on necessity rule.

Keywords

• artificial intelligence
• necessity
• common law
• self-driving cars