On the eve of the NATO summit in Ankara, Lieutenant General Guillaume Thomas, deputy commander of the Alliance’s Allied Air Command, admitted what has long been said behind closed doors: the air threat is evolving faster than the structures, processes and procurement cycles NATO was built around.
The summit itself brought Ukraine €70 billion in military support for 2026, with a commitment to at least the same level in 2027, recognition of Ukraine as a security donor to the North Atlantic Alliance and a fresh confirmation of the long-term Russian threat in the final declaration. On the sidelines of the summit, Ukraine signed drone deals with Estonia, the Netherlands and Denmark, and is preparing further ones with Canada and Finland; nine such agreements now exist in total. So Ukraine is moving from being a state in constant need of help to a partner with technology, knowledge and combat experience to share.
But there is a catch. All of these agreements are bilateral. We sign them quickly precisely because we are not spending time building new formats. What we lack as a state is a multilateral framework for cooperation. Ukraine is not, and will not soon become, a NATO member, so it cannot plug into the old collective security system the way it requires—or, for that matter, the way Europe's future security needs it to.
Europe itself, however, is increasingly thinking about a new architecture, one less dependent on the shifting moods in Washington.
What our drone deals actually contain
Of all our partners, the Nordic and Baltic states probably understand the reality of the Russian threat best. They need no convincing: their militaries, diplomats and defense ministries reckon with an attack on a NATO member within a couple of years and understand that the Alliance has yet to overhaul its doctrines and remains, in practice, wedded to the old maneuver and positional warfare of the Second World War. That is why the drone deals were struck with these countries first: they are prepared to solve the problem bilaterally, without waiting for consensus across the whole of NATO.
Ukraine's Ministry of Defense does not publish the texts of these agreements, but thanks to our Latvian colleagues' openness, we can get a sense of at least one of them—the one signed in Tallinn back on June 9. It contains a great deal: supplies of Ukrainian strike drones, ground robotic and maritime unmanned systems, financing for Ukrainian production, exchange of technology and experience and the buildup of integrated air defense—right up to a joint drone plant near the Latvian–Russian border. The partnership is set to run for ten years.
The specific forms and volumes of support, however, will be decided separately each year. These are frameworks that fix an intention to “keep moving in this direction.” The devil is in the details, or rather in the implementation: whether this turns into one-off projects or a powerful, systemic partnership depends on whether a mechanism emerges to turn a string of bilateral declarations into shared work.
The Patriot lesson: why a license doesn’t save the day
Donald Trump's announcement at the summit of readiness to grant Ukraine a license to produce Patriot interceptors is, without question, a tectonic shift and a diplomatic win for Ukraine. But upon closer inspection, there is not much to celebrate in the short term. Even in the best case, agreeing such contracts takes months—and Poland's experience shows it can take years. Then comes building the plant itself, which in Ukraine's case will also need to be defended daily from Russian strikes. And even once we have the license, we will run into the same wall every missile producer in the world is up against today—the components base.
The problem with producing missiles for the Patriot system is not a lack of production lines. A single PAC-3 missile is made of components from more than 400 suppliers, takes up to two years to produce, and some parts take even longer to source. Defense companies, even the largest ones, compete for components with the automotive industry—and lose because their orders amount to only a few percent of a supplier's annual output. At this pace, the missile simply doesn't get built, even as Ukraine's need is estimated at roughly 2,000 interceptors a year.
That raises a logical question: do we want to sink enormous investment and years of work into a plant that won't solve the problem here and now—and that risks become an obsolete technology by the time it launches? Almost every European NATO member has already concluded that current missiles no longer justify themselves, either on cost or on production speed, and has launched programs to find cheaper alternatives. Britain, in its new defense investment plan, has dropped further purchases of Storm Shadow missiles in favor of cheaper long-range alternatives, citing Ukraine's own experience directly.
And here is the key point: had Ukraine been more deeply embedded in Europe's security architecture, we would already be taking part in precisely these programs—the ones that will replace the PAC-3 and Storm Shadow. Instead, we again risk being left on the sidelines, collecting licenses for technology of yesteryear.
“Middle powers of the world, unite!”
Two days before the summit, the UK, Finland, Poland and the Netherlands released a joint statement on building a new Multilateral Defence Mechanism (MDM)—a model for pooling financing, procurement and demand aggregation for critical defense capabilities, due to launch by 2027 and expand to other countries as early as this autumn.
The logic by which states build alliances is changing before our eyes. Canadian Prime Minister Mark Carney and Finnish President Alexander Stubb recently spelled this out in a joint column for The Economist: the world order built after 1945 is falling apart, and its institutions, in their current form, cannot arrest that decline. The way out, as they see it, is a variable geometry of dynamic, purpose-built coalitions that form around shared interests rather than a single shared institution. Tellingly, the examples the two middle-power leaders cite: the coalition of the willing planning Ukraine's postwar security has already become the main forum for coordinating like-minded states and NATO; EU instruments such as SAFE let partners cut through red tape and build defense capabilities fast. Their conclusion: the question is no longer whether to build coalitions, but how fast we can build successful coalitions that scale. Their call—“middle powers of the world, unite”—is addressed, among others, to us.
This is exactly the frame of mind Ukraine needs to operate in. Not mourning the world as it was, and not waiting for membership in institutions that are themselves in crisis—but building for the world as it is. Carney and Stubb call their approach “values-based realism”: principled in its commitment to values, pragmatic in recognizing that progress comes through step-by-step dialogue. For Ukraine, this means not setting multilateral formats against NATO or EU membership, but building working coalitions now with those we align with on the essentials.
At the Ukraine Facility Platform, we have tried to sketch out how an architecture like this could work. The working name for the concept is PACS, Partnership Architecture for Collective Security. It is a model of targeted partnership, open to any interested European country, with a natural core made up of those who feel the threat most acutely—the Nordic-Baltic “eight,” the UK and Ukraine. Its underlying logic is a closed loop of joint capability-building. Military experts from Ukraine and its partners analyze battlefield lessons and translate them into a shared demand signal. On that basis come joint planning and joint procurement—through national funds, EU money, and donor initiatives, with economies of scale and lower prices.
Controlled technology transfer lets partners build their own production capacity, with Ukrainian demand forming the initial order book. And ongoing dialogue with industry, together with joint research and development should ensure that the technology doesn't go stale but keeps being updated at the pace the war demands. That means there are four tracks: political—aligning defense policies and removing legal and regulatory obstacles; military—turning Ukraine's lessons into shared doctrines and orders; technological—joint development from urgent front-line needs to long-term planning; and industrial—links between defense-industrial bases, joint production, and coordinated export controls.
There is also a purely political dimension to this construction that every Ukrainian politician needs to understand. Politicians in partner countries very often see the whole picture perfectly well, but they are constrained by public sentiment that our neighbor has spent decades, and continues to spend enormous money and effort, shaping. Too much pressure on friendly governments can cost them elections and bring to power the very people Russia is backing. A multilateral framework gives partners exactly the “cover” they are missing: a clear logic of mutual benefit, in which support for Ukraine is easy to explain to one's own voters, because it is no longer aid but shared security, shared factories, shared jobs and cheaper procurement for their own armies.
The main advantage of this approach is that it doesn’t require a general consensus—and that is exactly what makes it realistic. Start by assembling a core coalition of the willing, then scale up a process that already works. Bilateral agreements are the right and swift first step. But without a war strategy, a strategy for developing the defense industrial base and a strategy for multilateral cooperation, we risk once again being left on the sidelines of a process that will determine Europe's security for decades to come.