New Technologies in International Law / Tymofeyeva, Crhák et al.

fairness, and accountability at the core of AI governance; and combining existing norms, regulations, and standards into a framework for agile governance of AI. 218 Since there is no draft version of the Compact at the time of writing of this paper, it is questionable whether it will require States and other actors to respect, protect and fulfil the R2HE as a specifically mentioned human right. However, as it is common for UN documents on AI, only a vague reference to human rights is present in the documents. Furthermore, even if the Global Digital Compact is adopted, as such it will not be an international treaty, but only a non-binding document. 2.2 The European Union’s AI Act Probably the most active international organization on the issue of AI is the EU. The EU’s ambition is to be the leading actor in AI, aiming to boost research, industrial capacity and ensure protection of human rights and fundamental freedoms. Since 2018, the EU adopted several documents on AI, 219 however, in this paper we only focus on the draft of Artificial Intelligence Act 220 (hereinafter “AI Act”) and its regulation in the context of the R2HE. The extraterritorial application of the AI Act and its likely demonstration effect for policymakers means that the AI Act will have a range of implications for the development of AI regulation globally, as well as efforts to build international cooperation on AI. 221 The AI Act focuses exclusively on the high‑risk AI systems, which are defined as those that are part of a product falling under the EU product safety regulation or belong to a list of stand‑alone high‑risk AI systems laid down by the proposal, such as AI systems assessing the creditworthiness of individuals or used in the context of recruitment. 222 When it comes to the requirements for High-risk AI systems, Art. 9 states that a risk management system shall be established, implemented, documented and maintained in relation to high-risk AI systems and it shall consist of a continuous 219 See for instance: European Commission: Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee and the Committee of the Regions: Artificial Intelligence for Europe, 25 April 2018, COM/2018/237; European Commission: Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions – Fostering a European Approach to Artificial Intelligence, Brussels, 21 April 2021, COM(2021) 205 final; European Commission: Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee and the Committee of the Regions: Coordinated Plan on Artificial Intelligence, 7 December 2018, COM(2018) 795; European Commission: White Paper on Artificial Intelligence – A European approach to excellence and trust, Brussels, 19 February 2020, COM(2020) 65 final. 220 European Commission, Proposal for a Regulation of the European Parliament and of the Council on Artificial Intelligence (Artificial Intelligence Act), Brussels, 21 April 2021, 2021/0106(COD). 221 Meltzer J, Tielemans A, ‘The European Union AI Act: Next steps and issues for building international cooperation’ ( Global Economy and Development at Brookings , 1 June 2022) . 222 European Commission, Proposal for a Regulation of the European Parliament and of the Council on Artificial Intelligence (Artificial Intelligence Act), Brussels, 21 April 2021, 2021/0106(COD), Art. 6. 218 UN, Our Common Agenda Policy Brief 5: A Global Digital Compact – an Open, Free and Secure Digital Future for All, (2023) , p. 10.

66