Algorithmic discrimination; Legal Analysis

Using algorithms, especially artificial intelligence, to make different decisions, has raised many legal issues and questions. One of these issues is the discriminatory nature of algorithmic decisions. This discrimination may be against an individual, group or a class of people in the society. States have general obligation for elimination of any discrimination, including, Algorithmic  ones. Hence, identifying extensions and causes of algorithmic discrimination and providing solutions to deal with them has become one of the legal issues of nowadays. This article examines this question: what is algorithmic discrimination, what causes it, and how should it be addressed? To this end, while defining and categorizing algorithms and algorithmic decisions, the author has identifide two categories of causes (related to input data and related to design and performance) for algorithmic discrimination and by referring legal initiatives of leading countries regarding such discriminations as well as classifying various solutions. Finally, due to the lack of special rules to protect individuals against such decisions in Iran, the use of the strengths of all the solutions presented in the article has been suggested to fill this legal gap.