AI Liability Directive Timetable

Today, we want to give you some insight in our timeline and early ideas on this important legislative file that will complement the AI Act as last part of the EU AI strategy.

Many have asked Axel Voss and me about our position on the AI Liability Directive (AILD) as well as the next steps. It took us a while to have a first proper read of the relevant materials and to reflect on the way we want to continue working on this legislative file. We have concluded both points by now.

On 3 February, you will hear more about our initial findings but also about some of the questions we have for you (= 1st stakeholder consultation). With my post today, I want to give you a better idea of what we plan to do over the next 12 months so you can plan accordingly if the AILD is of interest to you.

As you know, Better Regulation and the 2016 Interinstitutional Agreement on Better Law Making are dear to us! Being the AILD rapporteur, Axel will therefore try to carry out a legislative process that is accessible, transparent, and empirical (evidence-based). It means, for instance, that three stakeholder consultations will give every interested party the chance to contribute, information on the legislative process will be regularly shared, and the relevant academic research as well as relevant use cases will be repeatedly taken into account during the political negotiations.

The attached timetable (1st slide) lists the key dates along the legislative process in the European Parliament. We plan to release the draft report in June with a deadline for AMs in July. The second half of the year will be reserved for the compromise negotiations with a planned vote in January 2026. Please note that all dates are subject to change as you cannot really plan ahead in Brussels 😅 and as they also depend on a potential alternative Impact Assessment. However, I will regularly update the timetable and let you know once a change is confirmed.

The attached infographic (2nd slide) summarises and illustrates our legislative approach.