The head of Palantir, Alex Karp, has spent years citing Ukraine as proof that software, data and decision speed can offset an opponent's advantage in scale. One caveat, without which quoting him would be dishonest: Karp is not an independent researcher. He sells precisely the systems he is describing. That does not mean he is wrong. It means what has to be checked is not his words but what is happening at the front.
The Pentagon says openly that cheap drones are the main technological break of the generation, and that American industry will take years to reach the necessary production rates. Ukrainian companies have entered the Drone Dominance programme. Ukrainian counter-drone specialists have worked in the Middle East, helping American forces among others.
A very strange historical moment: a country asking for missiles is already exporting the knowledge of how to survive a missile and drone war.
So — who owns, and who will own, the models trained on the Ukrainian war? Who registers the patents? Where does the intellectual property stay? Who receives licence income in ten years' time? What share does Ukraine get if a foreign corporation builds a global product on data obtained thanks to Ukrainian soldiers, engineers and society?
Combat data cannot simply be declared a commodity. Some of it is a state secret, some contains soldiers' personal data, some was created by Ukrainian companies or jointly with allies. Which is why Ukraine needs more than a demand for a share. It needs a clear legal architecture: classification of data, rules of access, licensing, allocation of intellectual rights and a fair return of the value created.
If Ukrainian energy, experience and data generate capitalisation exclusively outside the country, that becomes a new form of raw-material economy. We used to export grain, metal and ore. Now the risk is that we begin exporting engineers, telemetry and combat experience for free.
Ukraine acquired this technological knowledge at a terrible price. It should not only serve the security of allies; it should strengthen the security, the economy and the national capital of Ukraine itself. We have been through this with grain. Let us not repeat it with data. Partnership implies exchange. A resource is simply used.
That all sounds like theory, which is why two pieces of news are worth putting beside it. One you probably missed. The other you almost certainly did not.
In July, at the NATO summit in Ankara, President Trump told President Zelensky that the United States would give Ukraine a licence to manufacture Patriot interceptors: we will give you the right to make Patriots, we will show you how. And immediately added a phrase almost nobody noticed — we have not told the company about this yet.
Stop on the word licence. Not shells. Not money. Not guarantees. The right to produce exactly the type of asset this whole series is about. For the first time in four years Ukraine was promised not a thing but a right.
I would like to end that paragraph in triumph. I cannot, because what followed deserves a textbook of its own. On 11 July Reuters reported that Ukraine was unlikely to begin producing Patriots any time soon. On the 28th, after a meeting at the White House, Zelensky said an arrangement existed. On the 30th Trump told the Financial Times he was not sure. And on the 31st, at Camp David, he formulated it finally: we did not agree to this, handing over such technology is difficult, and those you give it to may one day turn it against you.
Three weeks. Four formulations. One right that appears and then disappears.
Indignation is the least useful available emotion here. I want you to see the mechanics. Even if the licence had been granted definitively, between a right and a capability lies an entire industry: tooling, hundreds of specialised suppliers, agreement with the prime contractors. Manufacturing interceptors of that class is limited even inside the United States.
There is the whole series in one example. A right can be obtained in a day at a summit and lost in a day at Camp David. The ability to use it is built over years, and it is built in advance, before the right appears. That is why all year I have been writing not about aid but about infrastructure. When the door finally opens, the person who walks in is not the one who asked loudest. It is the one who already has the plants, the engineers and the lawyers. A licence without an industrial base is an invitation to a banquet without cutlery.
And the second piece of news from the same week, the one that enraged Ukrainians so much that I could not go around it. On 1 August Trump stated that under the signed agreement on rare-earth resources the United States could enter Ukrainian subsoil at any time and take practically whatever it wanted, and that the deal was advantageous because the value of that access could exceed three hundred billion dollars.
The reaction was instant. The comments said: by the Constitution the subsoil belongs to the people; can a president hand over subsoil single-handedly; no security, no contracts. I read it, and I remembered that I had warned about this too, honestly enough.
Let us not discuss the quotation. Let us open the document.
The document is considerably narrower and more precise than the political quotation. The Reconstruction Fund agreement, signed on 30 April 2025, establishes this. Ukraine retains ownership of all subsoil and decides itself what is extracted. The fund is joint, fifty-fifty, with no dominant vote. It receives half the revenue only from new licences — existing projects, including Naftogaz and Ukrnafta, are excluded entirely. The agreement creates no debt obligations. Profit is not withdrawn for the first ten years.
The United States did receive a great deal: preferential access to critical minerals and the right to negotiate the purchase of output on market terms. That is a serious advantage. But it is not take whatever we want.
The difference between those two sentences is not a legal trifle. It is the difference between partnership and capitulation. And between a calm night and panic for millions of people.
I will say the part some people will not like, too. There is something in Trump's manner that is genuinely useful to us: he does not conceal his interest. He says aloud what others dress in diplomatic constructions about shared values. It is unpleasant to hear and it is useful, because illusions come off faster than any analysis.
But a characterisation of an agreement is not the agreement. The President of the United States describes the document the way he wants the American voter to see it: we will get our money back. That is his job. Ours is to read documents.
And here is the real conclusion, the one that makes me uneasy. The danger is not that something terrible is written in the treaty. The danger is that a document which protects us on paper depends on our own capacity to execute it, to insist on it, and to be needed.
We hold the right to our subsoil by the Constitution and by agreement. A political promise of a Patriot licence lived three weeks. The difference between those two cases shows everything.
Paper has force when institutions, capability and the readiness to defend what was agreed stand behind it. So that I am not misunderstood in the opposite direction: a right matters enormously. It simply does not work by itself. A document is not a shield. It is the drawing of a shield.
There is an old version of this. While Odysseus fights for ten years and travels home for ten more, suitors live in his house. They are not enemies and they do not storm the walls. Formally they behave decently: they court, they observe the customs, they wait for a decision. They simply eat his herds daily, dispose of his stores and decide the fate of his property. For years. The master is absent.
The most important thing in the poem is the ending, and it is not about revenge. Odysseus returns unrecognised, in rags. Nobody believes he is the master. There are no documents. There are no witnesses. He has no right to say this is mine; his words carry no weight. And then they bring out his bow, which nobody has touched in twenty years. The suitors try the string one after another and cannot. Not because they are weak — because it is not their bow. Odysseus takes it and strings it at the first attempt.
In the poem the master's right is confirmed not by paper but by the ability to use it. A twenty-first-century state cannot choose between those two things. It needs both: the right and the capability.