This title is reductive. AI ethics should ALSO be based on SOCIAL RIGHTS which are not human rights because social rights are collective whereas human rights are individual. Is AI going to improve the education of the nation within a project of guided self-learning? If not it could be taming, or even manipulating and conditioning. IS AI going to be a Brave New World project (each individual is in a category educated passively to a clearly-defined set of behaviors and rights, alpha, beta, etc.) or a real free individual responsible in society with RIGHTS and DUTIES? The superficially libertarian project is, in fact, a vast project of individuals locked up on and in themselves and thus manipulated out of any social dimension except the one conveyed by social friendliness within conformity, the inner-friendliness of the Swiss Guard in the Vatican.