A little more than two months after a green light from the European Parliament, it has received the final approval of this Act to be MADE by the Council of the European Union. Therefore, the legislative process of the law of artificial intelligence has been completed, but the implementation of it is not going to be immediate. For the full version of the documentation to be published, please refer to the notice of the press in the pages of the website of the institution. Here, the group's objectives pursued and the methods of application.
The green light on the writing of IT in Europe, and intelligent man
This is a set of regulatory and unique of its kind at a global level, defined by the stated objective of the harmonization of the rules for the AI, through an approach based on the assessment of the risks that are associated with these systems, using the principles of trust, transparency, and accountability. .
The IT was designed to ensure respect for the fundamental rights of the citizens for european, with no set, however, reduce to innovation, for as long as the innovations introduced by the sector and are consistent with what is expected of it. For example, it is possible to be developed, tested, and tested the technology inside of them, which are the names of the sandboxe (areas controlled environment) before you enter into the real world.
The management of the risk, and the risks are unacceptable.
The law etiketon the different types of artificial intelligence, in conjunction with the risks to which they lead, by applying the rules of the group, in accordance with the circumstances. Some have been the subject of a prohibition to be determined a priori, since they are exposed to the risks to be considered to be acceptable; this is the case of systems to manipulate cognitive-sjellës, and to those who make a point of social protection (see an article in the year of 2018, in connection with a project of this type started in China).
The use of the so-called policim predictor based on profiling and the use of biometric data to classify people according to criteria such as race, religion, or sexual orientation is also prohibited.
A new model of governance
Since the enactment of the Council of europe, there is also a goal of creating new bodies responsible for the application and update it to the potential of the AI is to the Board, and the AI is made up of representatives of the member states, and a forum for consultation, which will provide instructions on the necessary technique.
Heavy penalties on the offenders
For companies that are considered to violate the provisions of, and shall be punished to the grave. The amount will be paracaktohen, or estimated on the basis of the turnover of the previous year, as the case may be.
The time of the application
The next step is finding one that will go through publication in the Official Journal of the european union, of which it is expected already this week, and then it will enter into force 20 days later. The application is expected for the following two years, with a few exceptions, of the specific.
Discussion about this post