State of play on the Artificial Intelligence Act


The EU’s legislative proposal for the Artificial Intelligence (AI) Act is the world’s first attempt to regulate this booming technology. EU policymakers are strongly invested in the negotiations on the file. For instance, lengthy discussions within the European Parliament took place among the political families over which committees and groups would take the lead. More recently, members of the IMCO and LIBE Committees tabled over 3,000 amendments to Dragoş Tudorache (RE, Romania) and Brando Benifei’s (S&D, Italy) draft report. On the side of the Council, the upcoming Czech Presidency of the Council of the EU also expressed its willingness to achieve a General Approach in the next months.

In this article, we take a closer look at the legislative state of play on the file.

In the last few months, the AI Act has been high up on the agenda of both Council and Parliament. Despite the endeavours made by French negotiators, holding the rotating presidency of the Council of the EU in the first half of 2022, to reach a general compromise by the end of their mandate, no general approach could be agreed upon. Instead, French officials drafted a progress report on the negotiations, summarising the main points of contention among Member States as focussing on AI rules for law enforcement and on the extent of the regulatory framework for general purpose AI systems. This was closely followed by the publication of guidelines on facial recognition technology in the field of law enforcement by the European Data Protection Board (EDPB).

Within the European Parliament, negotiations on the file are jointly spearheaded by IMCO and LIBE Committees. To formulate the EP draft report, IMCO and LIBE Rapporteurs each one first informally drafted a separate report, respectively discussed within each Committee. The two reports were then combined into a single joint draft report on 20 April, which only included points that could be agreed upon by the two Rapporteurs. Divergences were initially left out but were formalised into amendments, which explains the consequent number of submissions. ITRE, Committee for opinion, also recently endorsed Eva Maydell’s (EPP, Bulgaria) opinion. The paper puts forward an “EU AI Regulatory Sandboxing Work Programme”, calling on the Commission to take on a facilitating role to support Member States to develop further or to start building in-house expertise to manage regulatory sandboxes. It also sets out the role of sandboxes as frameworks to enable firms’ compliance with this Regulation at the design stage of the AI system (‘compliance-by-design’).

Czech Presidency’s eagerness on the file

Czech officials announced their plans to hammer out the last pending issues on the AI Act. As already pointed out in France’s progress report, EU Member States are still concerned on the broadness of the scope, risking capturing innocuous AI systems. The Czech Republic laid out different scenarios on “four high-level outstanding issues which require more thorough discussion” which include, besides the definition of AI, the classification of high-risk AI systems and the governance structure, with a special question mark on the role of the Commission. According to the proposal text, the latter is empowered to lengthen and update the list of high-risk AI systems via delegated acts after the legislation is adopted, as well as endowed with powers to enforce the rules for these systems. The Presidency, which seems in line with the French approach to the text except for the language on national security exemption, is expected to draft and circulate a new compromise on 20 July. You can learn more about the Czech Presidency’s priorities in our article on the topic.

Finally, the liability rules for AI systems will be addressed separately in an Artificial Intelligence Directive, due for publication on 28 September.

Ecommerce Europe is closely following the discussions on the topic as AI will shape the future of the European economy. With the proposed Regulation, the EU stands a chance to become the global leader in AI technologies and to set the standard. If you wish to know more about our activities, please reach out to us at