Algorithmic forms of discrimination and the protection of people with disabilities

One of the multiple risks of the frequent use of artificial intelligence systems is algorithmic discrimination. The paper denounces the neutrality of algorithms, indicating that, sometimes, the digital revolution we are surveying generates new forms of discrimination. In this paper, we debate negative and positive aspects regarding AI in relation to discrimination on the workplace with a special regard to people with disability. AI systems should therefore not only aim to predict which employees have the right characteristics to be hired, but they should also fight against the algorithmic discrimination. In particular, it is suggested that humans control the transparency of algorithms, to prevent them from going beyond what are human potentialities, becoming an armor to some people’s responsibility and a tool to exclude minorities.