The idea of the law that will enter into force in the future is to regulate artificial intelligence so that it does not cause harm to society.
The European Parliament today approved the law on the regulation of artificial intelligence. The regulation, which is considered to be the world’s first comprehensive artificial intelligence legislation, is scheduled to enter into force possibly at the beginning of next year.
The idea of the law is to regulate artificial intelligence so that it does not cause harm to society. The regulation is stricter, the higher the risks are assessed.
Artificial intelligence is defined in four different categories based on risk.
Artificial intelligence applications that threaten fundamental rights are banned. An example of this is the prohibition of the use of facial recognition, except for certain serious crimes, such as the fight against terrorist attacks and the search for suspects of the most serious crimes. Social scoring systems made with the help of artificial intelligence, for example, are also prohibited.
The following are considered high-risk artificial intelligence systems, among others:
– systems related to critical infrastructure, such as traffic or water or electricity distribution,
– systems related to training, recruitment and health care and
– systems used in law enforcement, border control and, for example, processing visa applications.
Regulation of low-risk AI systems such as spam filters is light. According to the EU, most artificial intelligence services fall into the low risk category.
523 MEPs voted for the law and 49 voted against it. The member states are expected to approve the agreement in May.