MEPs on the European Parliament’s Committees on the Internal Market and Consumer Protection (IMCO) and Civil Liberties, Justice and Home Affairs (LIBE) - both of which are responsible for the dossier - are expected to decide on the European Parliament’s position on the ‘AI Act’ on 26 April. The text could then be put to the vote in plenary on 31 May.
MEPs managed to find compromises on most aspects, although their work is complicated by the rapid evolution of this technology. In particular, they agreed on the issue of the definition of AI systems, which, as requested by several Parliament political groups and a number of sectoral organisations (see EUROPE 13129/15), was aligned with the OECD definition.
The issue of so-called ‘high-risk’ AI systems has long been the subject of debate. At a meeting of the rapporteurs on 29 March - with three meetings scheduled between 22 March and 19 April (see EUROPE 13147/27) - the political groups decided that all users of ‘high risk’ systems would be required to carry out an impact assessment. This should, among other things, take into account the potential impact on fundamental rights.
Further clarification is needed on several aspects, such as the future Council for Artificial Intelligence, which is being advocated by several MEPs, including the rapporteur for the opinion of Parliament’s Committee on Legal Affairs, Axel Voss (EPP, German) (see EUROPE 13014/30) (see EUROPE 12928/24). While the idea seems to be accepted by the different political groups, its exact role and competences are still under discussion.
The interinstitutional negotiations can start as soon as Parliament has adopted its position, the EU Council having already adopted its position on 6 December (see EUROPE 13078/12). (Original version in French by Thomas Mangin)