The EU's First Defence Projects of Common Interest Ahead of the 2028 Budget

On 3 July the European Commission proposed that the Council designate the European Union's first five Defence Projects of Common Interest under the EDIP Regulation. The proposal, COM(2026) 358 final, is a procedural instrument, and it has been reported as one. Its annex is something else. For each of the five projects it lists every participating country, the objectives and the estimated financial size, which makes it the first published account of which member states are willing to build which capabilities together. The distribution of countries across the five projects does not track defence spending, geography or threat exposure in the way the surrounding rhetoric would predict.

The Commission opened a call for expression of interest on 16 February and received nine proposals by the 29 May deadline. It assessed them against the eight criteria in Article 35(2), with input from the External Action Service, the EU Military Staff and the European Defence Agency, and proposes designating five: DECODER for drones and counter-drone systems, Integrated Maritime and Seabed Defence, SPACE, Federated Integrated Air and Missile Defence, and Eastern Flank Watch.

Scale and participation move in opposite directions

The estimated sizes are in the annex and have gone largely unreported.

Project	Countries	Estimated size	Horizon
DECODER	26 MS, Norway, Ukraine	€3.5–5bn	2033
SPACE	15 MS, Norway	up to €24bn	2034
Eastern Flank Watch	13 MS, Norway, Ukraine	€60–100bn	2036
EU-FIAMD	14 MS, Norway, Ukraine	€55–80bn	2040
IMSD	15 MS, Norway, Ukraine	€43–72bn	2045

Total designated ambition runs from €185bn to €281bn.

The project with the most members is the one with the least money. DECODER has 26 member states and a financial size under 2 percent of the low-end total. Article 35(2)(e) sets the participation floor at four, so a count of 26 reflects choice, not a minimum met. The two projects carrying real money, Eastern Flank Watch and air and missile defence, have thirteen and fourteen members. Austria, Croatia, Hungary and Slovenia joined DECODER and nothing else — the one project a state could sign without accepting a territorial or industrial commitment. Breadth of participation here marks low cost and low exposure, not shared priority.

The map does not match the spending

Seven countries are in all five projects: Germany, the Netherlands, Poland, Finland, Sweden, Norway and Romania. France, Italy and Spain are in four. Greece and Cyprus are in four. Bulgaria is in two.

Leopard 2 in the latest A7V variant. Photograph: Bundeswehr/Marco Dorow 

Romania's position departs from the assumed hierarchy. It sits in every project, alongside the Union's industrial core and the Nordic states, while France, Italy and Spain are in four. Romania has raised combined defence, security and border-protection spending to 3.7 percent of GDP this year, against a NATO-definition target of 2.5 percent, joined NATO's Drone Edge procurement initiative, and became a founding member of the Defence, Security and Resilience Bank, which will site its Southern Flank regional office in Bucharest. The annex confirms the direction. Whatever criteria the Commission weighed, Romania satisfied them across the full set, which places it among the small group willing and able to commit to the capital-intensive projects rather than only the accessible one.

Cyprus sits in Eastern Flank Watch despite being an island in the eastern Mediterranean, far from the Russian border the project is named for. According to the annex, the project is "geographically focused on the Union Eastern border" but "open to extend its geographic area of interest." Whether it was drafted with Cyprus in mind or Cyprus joined and the clause followed, the eastern flank now works as a category rather than a location, and currently includes the eastern Mediterranean. Nicosia's year, the drone strike on RAF Akrotiri in March and its SAFE procurement soon after, explains the move.

Bulgaria is in DECODER and Integrated Maritime and Seabed Defence. It is absent from Eastern Flank Watch, the €60–100bn project running along the Union's eastern border. Romania, Poland, Slovakia, Czechia and the three Baltic states are in it. Bulgaria, a Black Sea state on that border, is not. The absence is notable because Eastern Flank Watch includes military mobility, counter-mobility and the expansion of production capacity. Bulgaria has sought €3.2bn from the EU's SAFE facility, holds an established ammunition base at Sopot, and has unused industrial land and a Black Sea coastline. On the distribution and threat-exposure criteria EDIP applies elsewhere, it is the kind of state these instruments were drafted to reach. The annex gives no reason, and none has come from Sofia. The position still has a cost, as these projects feed into the 2028 budget, and Article 35(19) makes late entry subject to the existing participants' consent.

Designation is a claim on the next budget

Reading designation as a funding decision misstates what the document does. The memorandum records that funding "may be provided by the budgetary envelope identified for that purpose in the Work Programme," that the decision is "without prejudice to any subsequent decision to award Union funding," and that participating states would be funded "in accordance with a dedicated procedure" it does not describe.

Article 3(4) of the EDIP Regulation allows up to 25 percent of the Programme's €1.2bn envelope to go to Article 35 actions. That ceiling is roughly €300m, set against designated ambition of up to €281bn. The projects are being given legal existence, a Council decision, a fixed list of participating states and an estimated size, three years before the Union negotiates the budget that could fund them at scale. The European Competitiveness Fund, beginning in 2028, is the plausible source of money on that order.

Designation therefore functions as a position rather than a payment. A project written into a Council implementing act, with member states named and a financial estimate attached, is harder to remove from a budget negotiation than a proposal that has not yet been formalised. Each state that joined a project has placed itself inside a structure that will argue for money in 2028, and the states in the two large projects have placed themselves inside the structures with the most to argue for.

A second consequence does not depend on funding. Article 35(16) allows the planning, construction and operation of production facilities tied to a designated project to be treated as an imperative reason of overriding public interest under the Habitats and Water Framework Directives, in the interests of defence under REACH, and of public health and safety under the Birds Directive. Designation carries permitting weight, which for a state seeking to host manufacturing may arrive sooner than any disbursement.

For readers approaching this as an investment map, Article 35(13) limits EDPCI funding to member states, associated countries and Structures for European Armament Programme composed of them. Companies are not recipients. They enter through contracts, to which Article 35(9) requires participating states to apply criteria equivalent to the eligibility rules in Article 9 of the Regulation.

Ursula von der Leyen, President of the European Commission. Photograph: Getty Images

What was not designated

Four proposals were not taken forward: European Robotic Combat Units for Land operations, Artillery Systems, Multi Domain Operation Platform systems, and European Defence Data Space for AI. The Commission found they did "not yet fully comply" with the Article 35(2) criteria, citing "the insufficient clarity with which their objectives are described" and "their overall level of maturity." All four may receive EDIP funding for development and be designated at a later call.

Two of the four are autonomy projects: ground robotic combat units and a defence data space for AI. Both were judged immature. Aerial drones, by contrast, were designated with 26 member states and a European Network of Drone Technology Hubs written into the project. The split follows what each field has behind it. The aerial base exists because of Ukraine: three years of wartime production gave Europe suppliers, tested systems and an agreed picture of the threat. Land autonomy and defence AI have no equivalent behind them, no comparable industry and no group of firms and agencies already building at scale. The decision describes the sector as it is, not as this year's strategy papers describe it, and it marks the two fields where Europe has declared an ambition it cannot yet build.

Outlook

The designations are not yet final. The Council must adopt the decision and could amend it first. The participation lists hold only as of the date of adoption. Article 35(19) lets a member state join a project later, but only with the approval of those already in it, which makes late entry a negotiation rather than a right.

Three things are worth watching. The first is whether the lists change before adoption, and in particular whether Bulgaria enters Eastern Flank Watch. The second is whether the four rejected proposals return at the next call, and which states take them up. This will show where the appetite for land autonomy and defence AI actually sits. The third is where the European Network of Drone Technology Hubs is built. Hubs that follow the distribution logic running through the rest of the Regulation give the eastern periphery real infrastructure.

The Commission must evaluate EDIP by 30 June 2027. That review coincides with the negotiation over the 2028 budget these projects feed into. States already attached to the large projects will have the strongest claim on that money.