Discrimination due to the automated processes of artificial intelligence systems is more than a hypothesis; it is a fact. The autonomy and opacity that characterise this technology give a particular dimension to the phenomenon, requiring new formulas in order to ensure adequate protection of the right to equality as an inalienable human and fundamental principle. The main objective of the contribution is to explain the crucial conflicts that AI transfers to the Legal Sciences regarding equality and non-discrimination. Consequently, as a second line of analysis, it is intended to analyse the solutions proposed in the regulatory field to mitigate the risk and cover the liability.