Introduction
The legal definition of Artificial Intelligence (AI) is a sine qua non for regulating AI. However, AI is a complex, dynamic, and evolving technology. Thus, a precise definition of AI is difficult. This paper reflects on the definition of AI as outlined in Article 3(1) of the European Artificial Intelligence Regulation 2024 (hereinafter referred to as the AI Act).[1] It adopts a doctrinal approach: by interpreting Article 3(1) and Recital 12 of the AI Act, which was recently supplemented by the European Commission’s Guidelines on the Definition of AI (issued on 6 February 2025).[2]
The concept of AI in the AI Act is vital for many reasons. First, the legal definition of AI determines the scope of the AI Act. Thus, it must be clear what constitutes AI. A concise definition of AI creates legal certainty and facilitates the effective enforcement of the AI Act. Second, understanding the definition of AI in the AI Act is crucial because it is the world's first comprehensive regulation for AI. It can potentially serve as a blueprint (due to the so-called Brussels Effect) for AI regulation in other parts of the world.[3] Moreover, the AI Act has an extra-territorial effect. For instance, the foreign entities offering AI systems or services in the internal market of the European Union (EU) must comply with the AI Act. Thus, the concept of AI in the AI Act relates to States or businesses outside the EU. It is not only a potential model for AI regulation but also a regulatory requirement for doing business within the EU.
The paper is designed in the following order. Part II of the paper outlines a brief legislative history of adopting the definition of AI. Part III explains key features of the adopted definition of AI in the AI Act, and finally, Part IV provides a conclusion.
[1] The final text of the AI Act was adopted by the EU parliament, (legislative resolution of 13 March 2024), and approved by the EU Council (14 May 2024), published in official journal on 12 July 2024, entered into force on August 1, 2024, but most of its provisions shall not be enforced until August 2, 2026. Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act) (Text with EEA relevance), OJ L, 2024/1689, 12.7.2024, ELI: http://data.europa.eu/eli/reg/2024/1689/oj
[2] ANNEX to the Communication to the Commission Approval of the content of the draft Communication from the Commission - Commission Guidelines on the definition of an artificial intelligence system established by Regulation (EU) 2024/1689 (AI Act). Brussels, 6.2.2025 C (2025) 924 final
[3]For example, the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, 2024, defines AI exactly in the same way as the EU AI Act. Article 2 of the Convention defines the ‘artificial intelligence system’ as “a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations or decisions that may influence physical or virtual environments. Different artificial intelligence systems vary in their levels of autonomy and adaptiveness after deployment.” The Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, 2024, adopted by the 133rd Session of the Committee of Ministers, (Strasbourg, 17 May 2024), Committee on Artificial Intelligence (CAI), CM(2024)52-final 17 May 2024.
