For months, the crisis around Ukraine has been discussed primarily through Russia’s intentions. Is Vladimir Putin preparing to invade, using military force to extract concessions, or deliberately maintaining enough ambiguity that nobody can distinguish between the two? This week has made that uncertainty harder rather than easier to resolve. Russia announced that some forces were returning to their bases, while the United States and NATO said they had seen no meaningful withdrawal and believed Russia had continued increasing its military presence around Ukraine. Fighting has intensified in the Donbas, where Ukrainian forces and Russian-backed separatists have faced each other since 2014. Russia and Belarus have also extended joint military exercises that were supposed to finish today.
Diplomacy is still happening, and Russia continues to deny that it intends to invade Ukraine. That fact matters. An attack has not happened, and writing as though one is inevitable would replace analysis with prediction. But after months of military mobilisation, increasingly urgent Western warnings and deteriorating conditions in eastern Ukraine, the probability of a major confrontation appears serious enough that another question deserves more attention: if Ukraine is threatened, what exactly did the international security system ever promise to do about it?
President Volodymyr Zelensky raised essentially this problem at the Munich Security Conference yesterday. His frustration is understandable because Ukraine occupies a peculiar position. It is neither neutral in any meaningful geopolitical sense nor protected by the collective-defence guarantee enjoyed by NATO members. It wants closer integration with the West, receives Western military assistance and has repeatedly been told that its sovereignty and territorial integrity matter. Yet Western governments have also made clear that their own soldiers will not automatically fight Russia on Ukraine’s behalf.
Ukraine therefore has substantial international support without an absolute security guarantee.
That distinction has become one of the most consequential facts in Europe.
It also has a history that begins with nuclear weapons.
When the Soviet Union collapsed in 1991, Ukraine became independent while large numbers of Soviet nuclear weapons remained on its territory. It consequently possessed what appeared, at least physically, to be one of the world’s largest nuclear arsenals. Ukraine subsequently agreed to relinquish those weapons and join the Nuclear Non-Proliferation Treaty as a non-nuclear-weapon state. In 1994, Ukraine, Russia, the United States and the United Kingdom signed the Budapest Memorandum, under which the signatories gave Ukraine security assurances concerning its sovereignty, borders and freedom from economic coercion.
Russia’s annexation of Crimea in 2014 already demonstrated that those assurances had severe limits.
This is where terminology becomes extremely important. The Budapest Memorandum provided security assurances, not the kind of automatic military guarantee contained in NATO’s Article 5. The United States and Britain did not promise that American and British forces would necessarily go to war if Ukraine were attacked. Russia committed to respect Ukraine’s independence, sovereignty and existing borders, a commitment that its seizure of Crimea plainly placed in direct conflict with.
Legally and diplomatically, those distinctions matter.
From the perspective of an ordinary Ukrainian, they may feel considerably less satisfying.
Ukraine surrendered nuclear weapons. Russia later took Ukrainian territory. The Western powers condemned the annexation and imposed sanctions but did not fight Russia to reverse it.
That sequence creates a difficult question for the entire international system: what is a security assurance worth if the state receiving it discovers that nobody is obliged to enforce it militarily?
The simplistic conclusion would be that Ukraine made a catastrophic mistake by giving up nuclear weapons. I think that argument needs more care than it usually receives. Ukraine did not possess a completely independent, operational nuclear deterrent in the same way that Britain or France does today. The weapons were inherited from the Soviet system, and maintaining an independent nuclear arsenal would have involved enormous technical, financial and diplomatic difficulties. The United States and Russia strongly supported denuclearisation, and Ukraine was attempting to establish itself internationally as a new state rather than begin independence through confrontation with both.
We should therefore avoid rewriting 1994 as though Ukrainian leaders possessed all the information available in 2022 and simply chose naively between “keep perfect nuclear deterrent” and “give it away for a piece of paper”.
They did not.
But perception matters in international politics, and the perception created by Ukraine’s experience may still be damaging.
The global non-proliferation system asks governments to accept that their security is better served without nuclear weapons. Countries are encouraged not to acquire them, and states already possessing them are supposed to pursue disarmament over time. That bargain becomes harder to defend if governments conclude that nuclear weapons are the one security guarantee major powers reliably respect.
Consider the incentives facing another country watching Ukraine.
It sees Saddam Hussein’s Iraq, which did not possess nuclear weapons when the United States invaded in 2003. It sees Muammar Gaddafi abandon Libya’s nuclear weapons programme and later lose power during a NATO-backed intervention. It sees North Korea develop nuclear weapons and remain extraordinarily difficult to coerce militarily. Then it sees Ukraine surrender inherited nuclear weapons and subsequently lose Crimea to a nuclear-armed neighbour.
These cases are radically different, and treating them as controlled experiments would be terrible analysis. Regime type, alliances, geography, military capability and international circumstances all differ. Nuclear weapons are not the only reason North Korea survives, nor did Libya’s abandonment of its programme mechanically cause Gaddafi’s overthrow.
Yet policymakers learn from imperfect examples.
The lesson they draw does not have to be academically correct to affect behaviour.
If enough governments conclude that nuclear weapons provide the ultimate insurance against foreign intervention, non-proliferation becomes much harder.
That may be one of the largest consequences of the Ukraine crisis even if no new war occurs.
Security guarantees therefore have value beyond the country receiving them. Their credibility affects future bargaining.
This is similar to alliance credibility. When the United States guarantees the security of Estonia through NATO, Estonia is not the only audience. Latvia, Lithuania, Poland, Russia and every other American ally observe whether Washington treats commitments seriously. A guarantee can deter because adversaries believe violating it will trigger a response.
But credibility creates its own danger. If countries make commitments everywhere because they fear appearing unreliable, they can eventually promise more than they are willing or able to defend.
A security guarantee is useful only if it is credible.
That means there should be fewer of them than political rhetoric sometimes implies.
Promising to “stand with” another country is cheap. Promising that your citizens will fight and potentially die for that country is fundamentally different.
Ukraine is discovering the distance between those statements.
This also explains why NATO membership has become so politically explosive. Ukraine sees membership as a route towards a credible collective-defence guarantee. Russia sees the same guarantee as a strategic threat because it would potentially place NATO’s military commitment directly along another part of Russia’s border.
The argument is therefore not really about bureaucratic membership in an international organisation. It is about the future distribution of military risk.
If Ukraine joins NATO, an attack on Ukraine could become an attack on the entire alliance. That makes Russia less able to use military coercion against Ukraine, which makes Ukraine safer. It also means a future Russia-Ukraine confrontation could more easily involve nuclear-armed NATO members, which may increase other forms of risk.
Security can be redistributed without being created from nothing.
One actor becomes safer because another actor’s expected cost of aggression rises.
That actor may then respond by trying to prevent the new arrangement before it exists.
This is one reason Putin’s demand for guarantees against Ukrainian NATO membership is so important. Moscow is effectively trying to preserve a strategic environment in which Ukraine does not receive a Western military guarantee.
From Russia’s perspective, this protects Russian security.
From Ukraine’s perspective, it preserves Ukrainian vulnerability.
The same proposed settlement therefore has opposite meanings depending on which side evaluates it.
This is where the language of “indivisible security”, used frequently in European diplomacy, becomes difficult. In principle, one country’s security should not come entirely at the expense of another’s. In practice, states disagree about which actions create legitimate security and which create unacceptable threats.
Russia says NATO expansion makes Russia less secure.
Eastern European countries say NATO expansion makes them safer precisely because they fear Russia.
If the solution is simply that NATO stops expanding whenever Russia feels threatened, Russian security takes priority over its neighbours’ choices.
If the solution is that NATO ignores every Russian security concern because applicant states are sovereign, the possibility of escalation remains.
International law helps determine rights.
It cannot eliminate the strategic consequences of exercising them.
The Donbas makes these problems more dangerous because violence can create facts faster than diplomats can interpret them. Shelling has intensified this week, and both sides accuse the other of provocations. Separatist leaders have announced evacuations of civilians towards Russia and ordered mobilisation. Ukraine denies planning an offensive.
This is precisely the kind of environment in which information becomes difficult to verify and politically consequential. An explosion, artillery strike or civilian casualty can be interpreted as evidence that the other side has begun escalation. Governments can then justify actions as responses rather than initiations.
That is why the public intelligence strategy I wrote about two weeks ago matters. The United States has repeatedly warned that Russia might use a fabricated or manipulated incident as justification for military action. Russia says Western governments are manufacturing hysteria.
The problem for everyone observing from outside is epistemic. We need to distinguish what happened from what governments say happened while those governments possess far more information than the public and strong incentives to shape interpretation.
This is where good journalism and independent verification become part of international security rather than simply commentary on it.
False information can change military decisions.
Accurate information can prevent miscalculation.
The ability to establish facts quickly therefore has strategic value.
There is another development this week that should concern anyone hoping the crisis is de-escalating: Russia and Belarus have extended their joint military exercises. Those exercises placed substantial Russian forces in Belarus, geographically much closer to Kyiv than many traditional Russian bases. The exercises were supposed to conclude today; Belarus has now said they will continue because of heightened tensions.
Again, an exercise is not an invasion.
But military geography affects options. Forces already deployed near potential operational areas can act more quickly than forces thousands of kilometres away. Capability does not prove intent, but capability changes the range of actions available to the decision-maker.
That is why Western governments continue treating troop movements as evidence even while Russia says it is merely exercising sovereign control over its forces.
Intentions are private.
Capabilities are observable.
Security policy therefore has to respond partly to the second when it cannot know the first.
This creates the risk of a spiral. Russia deploys forces because it says NATO threatens its security. NATO strengthens Eastern European deployments because Russia’s mobilisation appears threatening. Russia then cites those NATO movements as further evidence that its security concerns were justified.
Each response is defensive according to the actor making it and potentially offensive according to the actor observing it.
The security dilemma becomes dangerous when neither side possesses enough trust to interpret ambiguous behaviour benignly.
And there is almost no trust left here.
That may be why security guarantees matter so much. A credible guarantee changes behaviour because it reduces uncertainty. Estonia does not need to guess every morning whether America would care if it were attacked; NATO’s treaty establishes a framework for the response. Russia knows that framework exists too.
Ukraine has no equivalent certainty.
Its security exists in the space between international promises and national self-defence.
Zelensky’s frustration in Munich therefore deserves more than sympathy. It raises a legitimate policy problem for the West. If NATO is unwilling to admit Ukraine in the foreseeable future, and unwilling to fight Russia directly on Ukraine’s behalf, what sustainable security arrangement is actually being offered?
More weapons can increase the cost of attack.
Sanctions can increase the economic cost.
Diplomacy can reduce tensions.
None provides the certainty of collective defence.
Perhaps no Western government is willing to provide that certainty because the cost would be too high.
If so, that should be stated honestly.
Ambiguous promises can create moral comfort without strategic protection.
This is not an argument for NATO immediately admitting Ukraine. Doing so during an active confrontation involving disputed territory could create extraordinary escalation risks.
It is an argument for distinguishing aspiration from commitment.
Foreign policy becomes dangerous when rhetoric outruns willingness to act.
A country hearing “we stand with you” may interpret the phrase differently from the country saying it.
Ukraine’s experience since 1994 demonstrates why wording matters.
Assurance.
Guarantee.
Partnership.
Alliance.
These terms can sound interchangeable in political speeches.
They describe radically different distributions of obligation.
Perhaps the most mature form of foreign policy is not making the strongest promise available. It is making the strongest promise you are genuinely prepared to keep.
That is less inspiring.
It is also more credible.
As of tonight, diplomacy has not collapsed. Russia continues to deny that it plans an invasion. Western leaders continue searching for a settlement. There is still space for de-escalation, and we should resist turning serious risk into retrospective inevitability before anything happens.
But the crisis has already exposed a weakness in the European security order.
Ukraine sits between two systems. It is sovereign enough to choose a Western orientation but not protected enough to pursue that orientation without substantial risk. Russia lacks a legal right to dictate Ukraine’s future but possesses enough military power to make its preferences impossible to ignore. NATO insists that its door remains open but has not offered Ukraine the protection that walking through that door would eventually imply.
Everyone’s formal position can therefore remain consistent while the strategic situation becomes increasingly unstable.
That may be the problem.
International order does not survive on principles alone. It survives when principles are connected to institutions capable of enforcing them.
Ukraine was promised that its sovereignty and borders mattered.
The current crisis is testing what the word “matter” actually obliges anyone to do.
And for every country that has ever been asked to trust an international security assurance rather than acquire the means to defend itself independently, the answer will be watched very carefully.
