On 9 April 2025 the European Commission published the AI Continent Action Plan. If the EU AI Act answers the question of what must not be done, this action plan tries to answer where capability will come from.
Four main lines
The action plan unfolds along four directions: building large-scale AI data and compute infrastructure, improving data accessibility, driving AI adoption in strategic industries, and strengthening AI skills and talent.
The compute infrastructure strand continues the “AI factories” approach and goes further with a proposal for larger-scale compute clusters. The data accessibility strand points at Europe’s long-standing fragmentation problem: the scale advantage of the single market has not been fully realised in the data field.
Rebalancing regulation and competitiveness
Since the AI Act was passed there has been a continuing argument within Europe over whether leading with regulation drags down competitiveness. This action plan is the EU’s answer: keep the regulatory framework while filling in the three supply factors of compute, data and talent.
That orientation is consistent with the “simplify the compliance burden” thinking in the digital omnibus proposal. For companies it means compliance cost and capability supply in the EU market may change at the same time, and each needs to be assessed separately.
How it connects to the GPAI rules
The action plan does not change the timetable for the obligations on general-purpose AI models, nor does it replace the voluntary role of the GPAI code of practice. It deals with underlying capability, not with the compliance obligations themselves.
For model providers, two questions need to be kept apart: what must be done to comply, and what can be obtained in terms of infrastructure.
What companies exporting to the EU should watch
First, compute and data policy may create localisation opportunities. The construction of compute clusters and data spaces is usually accompanied by requirements for local deployment and local partnership. See our notes on cross-border data compliance.
Second, skills and talent policy affects the pace of landing. Companies setting up an entity in the EU and building a local team can look at how to plug into the relevant training and talent programmes.
Third, budget separately for compliance and for supply. See our note on the Parliament’s position on the AI Act omnibus; mixing regulatory compliance spending with infrastructure investment makes it easy to underestimate the real cost.
The EU is trying to show that regulation and competitiveness are not a zero-sum relationship.
This article is based on the European Commission’s AI Continent Action Plan and public information and is for reference only; it does not constitute legal advice.