The EU has issued a guide to the use of AI prohibited by the “Law on AI”

The European Union entered into force the first rules provided for by the Law on Artificial Intelligence (AI ACT), which is aimed at regulating the use of AI in accordance with the level of risk. This law prohibits certain scenarios of the application of technologies representing “unacceptable risk”, including manipulations using techniques affecting the subconscious, as well as social scoring, which can lead to discrimination of people.

Image source: Growtika/Unsplash

According to TechCrunch, the EU executive body on Tuesday published a guide explaining to the developers how to meet the necessary requirements. These recommendations relate to prohibitions on certain scenarios of the use of AI, the violation of which can lead to large fines – up to 7 % of the company’s global turnover or € 35 million (depending on what is greater).

«Recommendations are designed to ensure a uniform and efficient use of AI ACT throughout the European Union, ”the commission said in a statement. It is noted that the recommendations are not legally obligatory and the final interpretation and the application of the law will remain with regulators and courts.

The commission clarified that the document includes legal explanations and practical examples that will help interested parties understand and fulfill the requirements of the law. At the moment, the recommendations are published in the draft version and will be finally approved after transferring to all the EU official languages.

Although the “Act of AI” entered into force last year, its implementation continues. In the coming months and years, additional provisions of the law will come into force. At the same time, supervision of the implementation of the rules will be introduced in stages: until August 2, the EU member states must appoint bodies that will monitor compliance with the norms.