The AI Act: timeline (April update Vol 1)

Four additional Technical Meetings and three Shadows Meetings have been scheduled in the European Parliament for the next five weeks. Overall 14 compromise batches have been discussed at technical level covering the whole text, including many Recitals.

A political deal is very close due to 7/7 working weeks at technical level. There have been some compromise miracles such as with the AI definition, where we found a middle ground solution for two opposing views. The biggest obstacles towards a deal are: GPAI, AI Value Chain, governance, enforcement, prohibitions as well as ANNEX III.

Although it remains a possibility that the EPP Group cannot vote in favor of the AI Act (i.e. borderless prohibitions & high-risk areas, unbalanced solution on GPAI / AI Value Chain), a positive vote in late April becomes more and more likely.

Compared to the proposal of European Commission as well as the General Approach of the Council of the European Union, the report of the European Parliament could strike a much better balance between protection and promotion from / of AI. Numerous legal overlaps have been fixed, the conceptual empty EU AI approach has been filled with life, and realistic (!) risks mitigation measures have been added fundamental rights civil society

We are still at a crossroads: maybe the AI Act forces all AI developers out of EU as it is not implementable & enforceable ... or ... the new law achieves its goal of creating trustworthy AI made in Europe. Scenario 2 is now a real possibility.

Here you can find the full overview of what happened in the last months and is planned: