For the European Unionthere are risks that are “unacceptable”: for example, that artificial intelligence uses techniques to manipulate us, that it uses facial recognition in real time for private gain or that it takes advantage of the vulnerabilities of individuals: “The appearance, the race, ethnic groups, religion… political opinions etc”explains Óscar Corcho, professor of Artificial Intelligence at the Polytechnic University of Madrid.

Nor could it decide who has or does not have access to public services… “What school can your children go to, or what institute or university can they go to, beyond what we are using right now that is objective enough in general”, adds Óscar.

For this reason, from the European Union, they want human beings, and only human beings, to supervise the decisions made by the artificial intelligence: “The important thing is to know how to manage them well, to know to what extent we can use them, when they should be used and in what way.”

The decisions made by an AI will be divided into three levels: high, medium or low risk depending on how it affects the daily life of the population. Oscar, specify: “It will not change excessive things but it will put more order in aspects related to the appropriation and privacy of data.”

All these guidelines have to be approved in June in the European Parliament. Although They admit that they are having a lot of difficulty agreeing, because even those who study and train it do not know how they make any of the decisions, why they answer some things. “Therefore, in view of the application that this new regulation is proposing, they are going to be very difficult to name,” the expert clarifies.

Europe was already a pioneer in the regulation and protection of data, something they expect, repeat itself this year with the Artificial Intelligence legislation.