An "Important Project of Common European Interest" (IPCEI) is an EU state-aid instrument that differs from conventional research funding or single-country industrial policy. IPCEI allows several Member States to jointly fund cross-border projects in areas where the market alone cannot generate sufficient investment or overcome systemic risks — typically in cases where a technology is not yet market-ready, but its strategic importance for Europe's competitiveness and sovereignty is high.

IPCEI-AI is the application of this instrument in the field of artificial intelligence. Since early 2025 it has been coordinated by Germany's Federal Ministry for Economic Affairs and Energy (BMWE), with 18 EU Member States constructively engaged in the initiative.

Why this is conceptually significant

First, the initiative deliberately focuses not on general-purpose large language models (LLMs), but on sector-specific, industrial AI — models needed for autonomous manufacturing, AI-driven logistics and robotics, AI-assisted design, and the industrial metaverse. This is a deliberate departure from the US- and China-dominated race in general-purpose AI — Europe's advantage is being sought where it already has a resource: industrial sector data.

Second, the initiative applies a systemic, full-life-cycle approach — from data availability and sharing standards, to open and competitive foundation models, sector-specific solutions, AI operating infrastructure (AI-as-a-Service), and open-source availability for end users. It recognises that AI competitiveness does not depend solely on model development, but on the simultaneous development of the whole chain — data, infrastructure, deployment and governance.

Third, it is part of the EU's broader "AI continent" policy architecture, which identifies five critical areas — computing infrastructure, quality data, AI algorithms, talent development, and regulation. IPCEI-AI is intended to be combined with the already approved IPCEI on cloud computing technologies and the planned IPCEI on edge computing infrastructure into a single "8ra" initiative. This signals that the EU does not see AI infrastructure, cloud computing and computing capacity as separate policy areas, but as an interdependent technological cornerstone.

Europe's AI competitiveness is not being sought in model size, but in the sector-specific data that European industry already controls.

Latvia's dimension

Latvia already has an established national contact point (the Ministry of Economics), and the Investment and Development Agency of Latvia (LIAA) has launched a national call for expressions of interest for letters of support to participate in the IPCEI-AI initiative — following the same approach as in previous IPCEI waves (the Health IPCEI, microelectronics, cloud computing). This means that Latvian companies and research organisations already, in principle, have a route to participate in this Europe-wide initiative.

This is also an important context for the forthcoming National AI Strategy — Latvia already has a functioning cross-border instrument for industrial AI development, which the strategy can and should make use of, rather than building a parallel national solution from scratch.

Would you like to follow how IPCEI-AI and other EU AI initiatives are affecting Latvia's industrial and public sector?

Contact PPPA →

Source: BMWE — IPCEI Artificial Intelligence