Gender knowledge and Artificial Intelligence

Among the various types of biases that can be recognised in the behaviour of algorithms learning from data, gender-related biases assume particular importance in certain contexts, such as the Italian one, traditionally linked to a patriarchal vision of society. This becomes even more true considering the context of university education, where there is a strong under-representation of female students in STEM Faculties, and, particularly, in Computer Science Courses. After a brief review of gender biases reported in Machine Learning-based systems, the experience of the teaching “Gender Knowledge and Ethics in Artificial Intelligence” active since A.Y. 2021-22 at the School of Engineering of the University of Padova is presented.