A Review of the EU Regulation on Artificial Intelligence


A Review of the EU Regulation on Artificial Intelligence

Artificial intelligence is evolving rapidly across all industries and social activities, offering a wide range of economic and social benefits.

For this very reason, AI technology can also be a tool for manipulative social control practices and be highly abused. Such practices can constitute a violation of many regulations protecting fundamental rights, including respect for human dignity, freedom, equality, democracy, rule of law, non-discrimination, protection of personal data, respect for private life and children's rights. The European Commission has taken action to prevent serious violations that may be caused by artificial intelligence technology.

On April 21, 2021, the European Commission proposed a regulation on the use of artificial Intelligence (Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonized Rules on Artificial Intelligence (Artificial Intelligence Act) and Amending Certain Union Legislative Acts) and the member states (EU Council) announced their position on the bill on December 2022. The European Parliament held the final vote to state its official position on June 14, 2023, and the draft resolution was accepted by a vast majority. In the aftermath, the member states will come together and agree on a common text of the law.

Legal studies on artificial intelligence are continuing rapidly by the European Union (“EU”). The Draft Resolution (European Union Artificial Intelligence Law) approved by the European Parliament by a vast majority (the '’Resolution''), defines the artificial intelligence system and hence states the definitions of technical concepts such as emotion recognition system, biometric recognition system, training data, test data, etc.

In the period from 2021 to 2023, public opinion was sought on the European Union Artificial Intelligence Law, and in this process, the European Parliament has adopted the artificial intelligence definition used by the Organization for Economic Co-operation and Development ("OECD"). In this respect, the European Parliament has identified artificial intelligence systems in four main categories as “low risk, limited risk, high risk and unacceptable”.


  • Minimum risk artificial intelligence software such as video games will be allowed without restriction.
  • Limited-risk artificial intelligence, such as chatbots on banking or airline websites, can be used provided that they comply with transparency liabilities.
  • High-risk artificial intelligence such as self-driving vehicles and robotic surgery will be subject to strict controls.
  • The use of artificial intelligence for social scoring will be strictly prohibited. For example, systems that analyze the behavior of citizens and make evaluations on this basis will not be allowed.

The legal basis of the Resolution is Article 114 of the Treaty on the Functioning of the European Union (TFEU).

The Resolution aims to prevent the use of certain harmful AI applications, and impose certain restrictions to protect individuals regarding the processing of their personal data, especially in the use of biometric identification systems. The Resolution entails national public authorities to maintain confidence in the use of artificial intelligence and to strengthen enforcement mechanisms.

The desired end result is to create agreed-upon rules for the implementation of artificial intelligence algorithms. The ultimate goal of the Draft Resolution is to achieve a controlled progress by avoiding dangers such as extremism, misinformation and spurred conflict, which are the threats posed by artificial intelligence tools. The drafting of this Resolution tells us that it has become a priority to address the current challenges of artificial intelligence and to prevent the potential for its abuse by introducing legal/penal codes, while monitoring future risks. It is obvious that the limitations brought by the Resolution will have a great impact on the practices of today's tech companies that use algorithms and artificial intelligence. In the upcoming period, all eyes will be on the final text which the member states will come together and agree upon.






Similar Articles

September 2023 Using Artificial Intelligence in the Workplace: What are the Main Risks for Ethics?
February 2023 Mandatory Mediation Between Traders in case of Negative Declaratory Actions Arising From a Lease Contract of Commercial Nature