Artificial Intelligence 25.04.2026

On 17 April 2026, Inria, IMT and Fraunhofer submitted the final report of the Franco-German AI Executives’ Dialogue to Thomas Courbe (France) and Dr. Beate Baron (BMWK). The report is the direct blueprint for IPCEI AI and will strongly shape German and French industrial policy over the next 36 months. The COO-level read on the homework for mid-market executives is clear: companies that don’t engage with the political translation in 2026 will be chasing funding pots in 2027.

TL;DR: The 17 April report becomes the funding logic for 2027

• Inria, IMT and Fraunhofer delivered the final report of the Franco-German AI Executives’ Dialogue on 17 April 2026. Recipients: Thomas Courbe (French Ministry of Economy) and Dr. Beate Baron (BMWK).
• The report feeds into IPCEI AI (Important Project of Common European Interest), which from 2027 onwards will structure the funding logic for AI industry in Germany and France.
• Three core structural themes: sovereign compute sourcing paths, GAIA-X data spaces as a mandatory anchor, and an industry use-case roadmap with concrete pilot targets through 2028.
• For DACH mid-market companies, this creates a preparation obligation: any business wanting to compete within the IPCEI AI funding corridor must demonstrate GAIA-X-compliant data spaces and sovereign compute paths in 2026.
• Executives who have not strategically positioned this topic in their business roadmap by summer 2026 will lose one to two funding cycles.

What the 17 April report actually contains and what it aims for

The Franco-German AI Executives’ Dialogue launched in Berlin in January 2025, backed by the French Embassy in Germany and the research institutions Inria, IMT and Fraunhofer. Initial findings were presented in November 2025 at the “Adopt AI Summit”; the final report was officially handed over to the economic ministries of both countries on 17 April 2026. Structurally, the report follows a three-pillar framework: sovereign compute infrastructure, shared data spaces based on GAIA-X logic, and a coordinated industry use-case roadmap.

The political leverage of the report lies not in its standalone significance, but in its role as input for IPCEI AI — the ongoing Important Project of Common European Interest on artificial intelligence. IPCEI is the EU state-aid framework that allows member states to jointly fund strategic industrial projects without breaching competition law. Previous IPCEIs (microelectronics, hydrogen, batteries) have unlocked funding volumes running into the double-digit billions in Germany and France. IPCEI AI has the potential to land in the same ballpark.

For DACH mid-market companies, the question is therefore not whether the report’s content is convincing, but how quickly they align their own business roadmap with the upcoming funding conditions. Those who structurally commit to GAIA-X-compliant data spaces in 2026 will have the formal prerequisites to file IPCEI AI applications in 2027. Those who have bet on pure hyperscaler setups will need to play catch-up during the application phase.

Three Consequences for the Mid-Market Business Roadmap

Consequence 1: Data spaces built on GAIA-X logic will become a funding prerequisite. The report makes clear that IPCEI AI is not simply meant to subsidize compute infrastructure — it explicitly targets data spaces where multiple industry participants can exchange data securely and with full sovereignty. GAIA-X provides the architectural standards for exactly that. Mid-sized companies currently centering their data strategy on Snowflake or Databricks should evaluate in 2026 how GAIA-X compliance can be added as an additional layer. From an operational COO perspective, this is not an architectural break but an extension: the GAIA-X layer governs access and usage contracts, not the underlying storage technology.

Consequence 2: Sovereign compute paths need a defined allocation target. The report does not advocate abandoning hyperscalers — it calls for a deliberate mix of hyperscaler and sovereign compute. In practice, this means the roadmap needs a clear decision framework for every new AI initiative: does the workload land on hyperscaler, sovereign cloud, or on-premises compute? The share of sovereign workloads should hit a sector-relevant target value over a three-year horizon. Based on early political signals, serious funding applicants can expect a minimum sovereign compute share of 30 to 50 percent by 2027/2028.

Consequence 3: A use-case portfolio with clear industry anchoring. The report is emphatic that IPCEI AI is industrial application funding, not research funding. Mid-sized companies planning to file for grants need a well-structured use-case portfolio with sector relevance, KPI backing, and a scaling plan. Organizations that have nothing more than “we’re doing a bit of GenAI” on their 2026 roadmap will have nothing fundable to show in 2027. From advisory practice, building a solid use-case portfolio takes three to five months. Anyone who does not start by summer 2026 will be late for the 2027 funding cycle.

What Mid-Market Executives Should Decide Before Summer 2026

Drawing on scaling experience in the DACH mid-market, four board-level decisions can be identified that should be made by the end of summer 2026 — so that the funding logic of 2027 can be used to its full potential.

First decision: Who on the board owns the IPCEI roadmap? A COO mandate with a clear line into the CDO and CFO is the 2026 standard, because the topic bridges industrial operations, data strategy, and funding finance. Pure CIO ownership is too narrow — the business linkages are missing. Second decision: Which use cases should be eligible for funding in 2027? The answer requires sector expertise, KPI modelling, and a realistic cost plan — not just a list.

Third decision: Which data spaces is the company building with partners? IPCEI AI specifically promotes multi-stakeholder data spaces where value chains share their data. Mid-market boards should spend the next 90 days assessing which suppliers and customers could jointly build a GAIA-X-compliant data space. Fourth decision: What funding corridor is actually plausible? Pre-screening by an experienced funding advisor with IPCEI track record is time well spent in 2026 — corridors look very different depending on the sector.

Executives who make these four decisions at their summer board retreat will have the structural prerequisites to actively draw on IPCEI AI funding in 2027. Those who don’t will be reading the announcements in 2027 and realising the preparatory groundwork is missing. From a COO perspective, that is the most expensive 90-day decision delay there is — and entirely avoidable.

How the Report Shifts Negotiating Power with Suppliers

One aspect that received little attention in the initial commentary on the report is its effect on supplier negotiations. If IPCEI AI starts funding sovereign compute paths from 2027 onwards, it simultaneously alters the negotiating position vis-à-vis both hyperscaler providers and sovereign cloud operators. Hyperscalers will increasingly knock on the door in 2026 with “Sovereign Cloud” offerings that emphasise European presence without actually delivering architectural sovereignty. Sovereign providers — T-Systems, OVHcloud, IONOS, Schwarz Digits — will arrive with bundled offerings that package compute, stack aggregation, and consulting together. In this environment, mid-market boards should have a written supplier strategy that explicitly weighs hyperscaler and sovereign paths against each other. Setting up that supplier playbook in the next 90 days avoids ad-hoc decisions made under sales pressure.

What Sets the Mid-Market CEO Perspective Apart from the Corporate View

In mid-sized companies, the structural prerequisites for IPCEI AI funding applications are typically quite different from those in DAX corporations. Mid-market firms generally have smaller funding teams — or no dedicated grant management function at all. Applications are often filed by external consultants, but preparing the substantive content remains an internal responsibility. That means: any mid-market company that wants to successfully apply for funding in 2027 must have at least one person in-house by summer 2026 who can independently manage the application process on a content level. Pure external consulting without internal substance will not hold up during the validation phase.

Experience from predecessor IPCEIs — Microelectronics, Hydrogen, Batteries — reveals four patterns that recur in the mid-market. First: mid-market applicants who enter jointly with two or three larger partners from their value chain achieve significantly higher success rates than solo applications. Second: industry associations play a bigger role than commonly assumed, both in the pre-selection of projects and in political accompaniment through the Brussels process. Third: structured ties to research institutions — Fraunhofer, Inria, KIT, TUM — measurably increase formal credibility and therefore funding probability. Fourth: a clear commercialisation roadmap for the funded solution has been a central evaluation factor since the Battery IPCEI, and will play the same central role in IPCEI AI.

What the First 90 Days Should Concretely Look Like

From an operational COO perspective, preparation can be broken into three clearly defined 30-day blocks. The first 30 days are about situational awareness: which internal AI initiatives are already running, which partners are potential candidates, which funding advisors have IPCEI experience, which industry associations are suited as a preparation platform. The second 30 days focus on architectural prerequisites: GAIA-X conformity check, sovereign compute share, data classification. The third 30 days are about the application framework: use-case definition, consortium formation, commercialisation roadmap.

Companies that work through this 90-day framework consistently from May to July 2026 will have a first internal application draft ready for board sign-off at the summer retreat in August. Those who start in September lose the strategic preparation time that serious funding applications demonstrably require — typically twelve to eighteen months. From a COO standpoint, this time window is the single most underestimated lever available to mid-market companies in 2026.

Which Questions Should Structure the 2026 Board Retreat

Scaling practice across the DACH mid-market distils six guiding questions that belong on the agenda of any serious board retreat dedicated to IPCEI preparation. Which two or three strategic AI use cases will we pursue with partners rather than building alone? Which industry alliance or association is the natural anchor point for our application position? Which research institution is relevant as a co-applicant, and how long does onboarding take? What internal capacity are we freeing up in 2026 for application preparation? Which external funding advisor has IPCEI experience in our specific sector? And what timeline and financial risks are we willing to absorb if the application does not clear the first round?

Any board that can formulate clear answers to all six questions has positioned itself strategically. One that responds to three or more with “still to be clarified” has homework for the next 90 days that cannot be delegated. IPCEI AI will become funding reality for AI in the DACH mid-market in 2027. The boards that prepare in a structured way now will extract the most durable advantage from the funding cycle — because by 2027 they will no longer be catching up on structural groundwork, but entering the funding corridor in a position to apply immediately. That preparatory time cannot be delegated and cannot be compressed; it must be built into the 2026 board calendar, running in parallel with ongoing operational priorities.

Frequently Asked Questions

What is IPCEI AI and when does it actually launch?

IPCEI AI is the ongoing EU state aid project supporting strategic AI industrial initiatives in Germany, France, and other member states. The first funding announcements are expected in 2027, subject to final notification by the European Commission. The Franco-German AI Executives’ Dialogue feeds into the programme definition as the central industry consultation.

Do we need to adopt GAIA-X to be eligible for IPCEI AI funding?

The report points strongly in that direction, though a final commitment in the funding programme has yet to be made. Companies that strategically embed GAIA-X compliance into their data architecture in 2026 will have the structural prerequisites in place for the 2027 application phase. Those that don’t risk a significant catch-up effort that can easily cost six to nine months.

Which sectors are particularly eligible for funding?

Indications from predecessor IPCEIs point to industrial use cases with clear strategic value: mechanical engineering, mobility, energy, healthcare, agriculture, and defence. Consumer goods and pure Software-as-a-Service applications without an industrial link have historically been less relevant for funding. The final sector list will be published alongside the programme announcement in 2027.

How much funding volume can realistically be expected?

Predecessor IPCEIs — covering microelectronics, hydrogen, and batteries — jointly reached funding volumes of 15 to 25 billion euros in Germany and France combined. IPCEI AI is expected to land in a similar range. For individual mid-market projects, realistic funding amounts run from 5 to 60 million euros depending on sector and scaling plan, with leading-edge projects exceeding that substantially.

How does IPCEI AI differ from the AI Act?

The AI Act governs AI compliance and risk classifications; IPCEI AI provides financial support for AI industrial initiatives. Both logics run in parallel and need to be considered together in any corporate roadmap. A pure compliance stance without a funding strategy leaves significant leverage on the table in 2026; a pure funding strategy without compliance preparation carries legal risk.

What do we do if we can’t reach 30 to 50 percent sovereign compute load?

These thresholds are political indications, not hard barriers. Applicants should state their current share transparently in the application and outline a migration path toward higher percentages. In practice, applicants with a clear migration roadmap are typically fundable even when the starting share is low. What matters is a credible trajectory across the funding period.

Related Reading on MyBusinessFuture

• A mid-market operations perspective on three takeaways from the report
• Deeper look at the scale gap according to the Infor Enterprise AI Index
• Background on the structural question behind AI failure rates

Cover image source: Pexels / Angel Bena (px:10142678)