The controversial draft Directive on the liability of artificial intelligence systems (the ‘AI Liability Directive’) is one of the legislative proposals due to be “withdrawn” by the European Commission, according to its 2025 work programme published on the evening of Tuesday 11 February (see other news).
This Directive was supposed to incorporate a civil liability system for damage created by AI systems and was intended to close legal “loopholes” in terms of legal proceedings, particularly where the onus is on the aggrieved consumer to prove the causal link between the AI’s action and the damage suffered.
German MEP Axel Voss (EPP), rapporteur for the text in the European Parliament and its Committee on Legal Affairs, described the planned withdrawal of the proposal as a “strategic error”. In a response to the media outlet Euronews, the MEP stated that this withdrawal would result in “a fragmented patchwork of 27 different national legal systems, suffocating European AI start-ups and SMEs”.
When asked about the precise reason for withdrawing this Directive, in the wake of the US Vice-President’s outright attack on the EU’s excessive regulation (see EUROPE 13577/12), Commissioner Maroš Šefčovič simply stated that the Commission was taking note of the “lack of progress” in the legislative process and that this withdrawal announcement was also a way of obtaining feedback from the co-legislators on the issue.
“We are waiting for them to tell us whether they really want to work on this legislation or whether they think the issue should be approached differently”, he said at the press conference in Strasbourg on the morning of Wednesday 12 February.
The AI Liability Directive was proposed alongside the AI Regulation in 2022, but, according to the Commission, there is “no foreseeable agreement” in sight for the proposal. (Original version in French by Isalia Stieffatre)