This translation was generated using AI and edited lightly. Read the original in French here.
Atlantico — With the multiplication of hybrid incidents attributed to Russia, experts have pointed to NATO’s “low profile.” Between expressions of institutional solidarity, diplomatic measures taken individually by states, and the rearmament undertaken by some countries, where does the real response lie today?
Frédéric Dulin — Clausewitz articulated a rule that the current sequence of events is confirming point by point. Whoever uses force without restraint against an adversary who refrains from doing so as a matter of principle gains an advantage that nothing offsets. That is exactly what has been happening on the continent for the past eighteen months, and there is no reason to describe it otherwise.
The real response today lies nowhere, and that is the reality we must acknowledge before anything else. The allied system has three levels, none of which imposes a cost on the party taking action.
The first is declaratory. On September 1, Berlin formally attributed the attempted attack in Leipzig to Russia. In early August, a drone carrying explosives had been discovered in the cargo area near Ukrainian cargo aircraft, on a platform essential to German and allied logistics. Ursula von der Leyen denounced an attack carried out on European Union territory by Russian agents using military equipment. Kaja Kallas described it as bearing the hallmarks of state-sponsored terrorism. Mark Rutte invoked Alliance unity from Brussels. Paris summoned the Russian chargé d’affaires. None of these responses contained a coercive element. They characterize the act; they do not punish it. Europe produced a legal case file where it should have produced a bill.
The second level is diplomatic and national. Germany will close its last Russian consulate, in Bonn, on September 18, along with the Russian House in Berlin. It is increasing pressure on the shadow fleet and proposing that additional Russian nationals be added to European sanctions lists. Romania had expelled the Russian consul; Moldova declined to do so, explicitly arguing that it would not stop Russia.
Two lessons follow. Each state bears the political cost of firmness alone, which mechanically discourages initiative. And the supply of diplomatic reprisals is finite. Closing your last consulate amounts to burning a rung of the ladder without building a new one, weakening Germany’s position at the next stage rather than strengthening it.
The third level is capability, and it is the only one with real substance. Eastern Sentry, launched in September 2025 after roughly twenty Russian drones entered Polish airspace and Article 4 was activated; Baltic Sentry, to protect undersea infrastructure; the establishment in June 2026 of a ninth multinational battlegroup in Finland under Swedish leadership; the expansion of battalion-sized formations to brigade-sized formations on the eastern flank; all of it supported by the 5 percent of GDP target agreed at The Hague summit for 2035.
This is a real, costly, structural effort. But its effects are expected in ten years, while incidents are occurring every few weeks.
The situation can be summarized in one simple sentence. The Alliance has substantially improved its ability to identify what is happening but has not developed mechanisms capable of imposing proportionate and effective punishment. Intelligence identifies. Judicial authorities trace the networks. Diplomacy protests. Defense hardens itself. At no point in this chain does anyone impose a cost.
This asymmetry is not accidental; it is structural. NATO was designed around Article 5—that is, around a clear act of aggression—and leaves everything below that threshold to national resilience, which is the responsibility of individual member states. Hybrid action therefore occupies an institutional blind spot that is perfectly visible from Brussels, and equally visible from Moscow.
Add to that the consensus rule, which turns every coercive measure into a negotiation among thirty-two countries, at a time when gradual American disengagement is depriving Europeans of the umbrella under which they once took risks.
The result is that Moscow reads the system exactly as it is: an architecture of protection without a mechanism for retaliation.
The operational conclusion the Kremlin draws from this is entirely rational: continue, and move up one level. In its place, any strategist would do exactly the same thing.
Tsiporah Fried — I think we first need to shift the question slightly. The issue is not so much whether NATO is responding as whether we have correctly defined what we are facing.
The term “hybrid warfare” has become extremely convenient. It allows us to group cyberattacks, sabotage, influence operations, drone incursions, and attacks on critical or undersea infrastructure under one label. But it also carries a risk: it can lead us to view each of these events as a separate incident requiring a specific and proportionate response.
The adversary does not necessarily think that way.
That may be where the fundamental asymmetry between Russia and us lies today. It is not only in the means being used; it is in our respective conceptions of the confrontation.
We Europeans still largely think in categories: peace, competition, crisis, then war. And we associate different instruments with each one. A cyberattack belongs to cybersecurity. Sabotage belongs to domestic security or intelligence. An information operation belongs to countering foreign interference. An airspace violation belongs to defense.
That division makes perfect sense from the standpoint of our institutions. Strategically, it makes much less sense.
Russia, by contrast, may view all these instruments as components of a single confrontation.
NATO itself now recognizes that significant cyberattacks and other hybrid attacks may, depending on their nature and scale, fall under Article 5. But the problem is less legal than strategic: who decides that a set of dispersed actions constitutes an attack against the Alliance, and at what point?
That is why I would be more cautious about characterizing NATO’s response as simply “low-profile.” The Alliance is not inactive. It has substantially reinforced its eastern flank, its defense plans, its cyber capabilities, and its resilience. Europeans themselves are rearming.
But we have developed our ability to withstand hybrid attacks more than our ability to deter their use. We are increasingly good at detecting, attributing, and absorbing these attacks. We are still poor at altering the calculations of the actor conducting them.
The real question is therefore no longer: are we being attacked? We already are, in multiple forms.
The question is: do we still view these actions as a succession of security incidents, or are we beginning to see them as components of a strategic confrontation?
Our weakness may be precisely that we look separately at tools Russia uses as instruments of a single strategy.
Atlantico — Can we say that Russia is proceeding through carefully calculated stages, with each incident remaining just below the threshold that would force NATO to respond collectively, while individual states rely on isolated measures rather than a structured defense effort? Is this the boiling-frog mechanism at work?
Frédéric Dulin — The step-by-step mechanism is real, but it is not quantitative. It is categorical, and Europeans themselves validate it at every stage.
Thomas Schelling established that the strength of a threshold lies in its visible qualitative character, not in the intensity of the damage. A threshold holds because it is clear, not because it is high.
Moscow therefore does not calibrate the power of its actions; it calibrates their legal classification.
The final operation in Leipzig involved so-called disposable agents, recruited outside the intelligence services and acting on orders from Russian state authorities. No uniforms, no claim of responsibility, no visible state apparatus on the ground. Nothing therefore falls clearly into the category of an armed attack, the only category that triggers a collective mechanism.
Russia is not testing Article 5. It is avoiding it.
The treaty is clear about the trigger: an armed attack. But it does not define what exactly constitutes one, nor does it specify the nature of the assistance that must follow, since each ally remains free to take whatever action it deems necessary.
The uncertainty therefore lies not in the threshold itself, but in how an act is classified before the threshold and how the response is defined afterward. Moscow operates methodically in this double gap.
Estonian Defense Minister Hanno Pevkur put it in his own way when he observed that Russia is trying to determine how far it can go without provoking a collective response.
This approach also reflects a doctrinal precedent commentators have somewhat misleadingly called the Gerasimov doctrine. The 2013 article began with an observation about what its author regarded as the American method of regime change seen during the Arab Spring. But he immediately drew a prescription for the Russian armed forces: adapt their forms of warfare to a new kind of conflict in which nonmilitary means outweigh military means, in a ratio the chief of the General Staff at the time estimated at four to one.
The commentators’ error concerns the origin of the method, not Moscow’s adoption of it.
The essential point remains the Russian belief that confrontation is a permanent continuum rather than an alternation between peace and war. Europeans, meanwhile, have never stopped thinking of peace and war as mutually exclusive conditions, leaving them poorly equipped to deal with something that is neither one nor the other.
The choice of targets also deserves close attention because it reveals more than a simple desire to cause disruption.
Leipzig is a logistical hub for support to Ukraine and the European base of Antonov Airlines. German domestic intelligence had already attributed to Moscow, in July, the burning of a package at a DHL logistics center located on the same site.
Drone activity forced the temporary suspension of operations at Munich airport. Undersea cables in the Baltic have repeatedly been damaged.
This is not diffuse harassment. It is an interdiction campaign targeting Europe’s lines of communication, conducted with an extremely low signature.
The target is not public opinion. It is the flow of goods and support.
The boiling-frog metaphor should therefore be discarded. It presupposes a failure of perception and offers a convenient excuse. Europe sees what is happening and documents it quite effectively.
Biological traces recovered from the drone, from a can filled with explosives, and from an antenna found near the airport allowed German criminal police, by cross-checking Lithuanian and Finnish databases, to identify two Russian passport holders in less than a month.
The attribution chain works remarkably well.
What is missing is not awareness of the danger, but the decision and the method required to impose effective retaliation.
Europe is not asleep. It is informed, it has documentation, and it is immobile.
The more accurate formulation is not gradual boiling, but a structural reversal of deterrence. The aggressor deters the defender.
Every Russian move is followed not by retaliation but by a European calculation about escalation risks, effectively giving the Kremlin control over the tempo and the choice of battlefield.
Each new stage therefore holds not because it has been cleverly calibrated, but because the absence of punishment validates it and authorizes the next one.
An incident that goes unpunished becomes a norm. A norm that goes uncontested becomes an acquired right.
That is why Vladimir Putin will not stop. Not because he is insensitive to risk, but because no risk has ever been presented to him.
European restraint is not prudence. It is a decision, and that decision will carry a price.
What comes next is already discernible. US intelligence speaks of the possibility of a limited attack against an Alliance member in a window running from autumn 2026 to 2029, while European intelligence services say they see no preparations for an attack currently under way.
Both propositions can be true simultaneously, and that is precisely the problem.
A limited action against an ally does not require lengthy and visible preparations. It requires only the conviction that it will go unpunished.
Europe itself is creating that conviction, incident by incident, communiqué by communiqué.
It is methodically creating the conditions for the attack it fears.
There is also a political dimension, which is the campaign’s true objective. Interference is increasing as decisive elections approach in Germany, France, Italy, Poland, and Spain, and at least three candidates in the French presidential election have already been targeted by disinformation operations.
The appearance since March 2026 of a supposed “People’s Republic of Narva” on Russian-language social networks follows the same logic of narrative preparation seen in the Donbas before 2014.
The objective is not conquest. It is to create doubt in advance so that allied decision-making will be paralyzed when the moment comes.
Tsiporah Fried — The boiling-frog metaphor is interesting, but I think it needs to be taken a little further.
Russian strategy is often presented as a succession of calculated provocations: Moscow gradually increases the pressure while always remaining just below the threshold likely to trigger a NATO military response, particularly under Article 5.
There is obviously some truth to that analysis. Russia actively manages escalation. It tests our reactions, measures our tolerance thresholds, exploits our divisions, and adapts its level of coercion accordingly.
But simply saying that Vladimir Putin wants to “stay below the threshold” is, paradoxically, still a very Western way of looking at Russian strategy.
It assumes Moscow shares our conception of a relatively clear boundary between peace and war and is simply trying not to cross it.
I think reality is more complex—and therefore more dangerous.
Russia does not necessarily consider itself to be at peace with Europe today.
Cyberattacks, sabotage, economic coercion, information manipulation, clandestine operations, and military pressure can all form part of a continuum of confrontation.
And that is precisely where the boiling-frog metaphor becomes relevant.
The problem is not only that the temperature is rising. It is that we continue measuring it attack by attack while the adversary thinks in terms of campaigns.
A cyberattack today, sabotage tomorrow, a drone incursion a few weeks later: each of these events, considered in isolation, may appear insufficient to justify a major response.
We therefore gradually become accustomed to a level of confrontation that probably would have seemed extraordinarily high just ten years ago.
It is this process of normalization that is dangerous. It allows Russia to test us and gradually raise the thresholds of those tests.
The risk is therefore not that we wake up one morning and suddenly discover that we are at war.
The risk is that we are already engaged in a strategic confrontation without having adapted our doctrine and instruments to that reality.
I would also mention a paradox that has emerged with the beginning of the election campaign: while we must take Russian interference very seriously, we must also guard against a kind of interference paranoia.
If every political crisis, every controversy, or every institutional dysfunction is automatically interpreted as the product of Russian manipulation, we end up attributing to Moscow a degree of power it probably does not possess.
Underestimating Russia is a strategic error. Attributing to it the ability to produce all our political crises is another.
Atlantico — If Europe is trapped by this gradual increase in pressure, where is the principal weakness in NATO’s conventional deterrence doctrine?
Frédéric Dulin — Russia is not exploiting one European weakness; it is exploiting a convergence of weaknesses.
Thirty years of peace dividends produced material weakness: ammunition stocks pared to the minimum and industrial bases reduced to token production lines.
The American repositioning created uncertainty about the security guarantee: a January 2026 strategic document refocused US forces on the homeland and the Indo-Pacific, followed by reductions in capabilities committed to allied defense plans and a complete review of the military presence launched in late July.
And Europe’s decision-making architecture created slowness.
A continent can be materially weak and still remain deterrent if it makes decisions quickly and strikes hard.
Europe is weak, dependent, and slow.
It is this combination, rather than any one factor in isolation, that makes the exercise attractive to Moscow.
The principal weakness does not lie in conventional deterrence doctrine. That doctrine works, and that is precisely why the attack has shifted elsewhere.
The problem is the complete absence of doctrine below the threshold, combined with an implicit strategic choice that has never been openly debated: the choice of the shield.
Six breakdowns accumulate and reinforce one another.
The first is legal. The allied architecture is binary.
Article 4 produces consultation. Article 5 remains unusable for an explosive device placed in a cargo zone or a cable severed by an anchor.
The space between the two is the adversary’s true operational environment, and doctrinally that space is empty.
International law nevertheless offers an available instrument: the theory of accumulation of events, under which a series of low-intensity acts may be treated as a single armed aggression when they form part of a common design.
No ally has undertaken the legal and political work necessary to make that concept operational.
The second weakness is the heterogeneity of rules of engagement along the flank.
In September 2025, Dutch F-35s and Polish F-16s destroyed several drones over Poland, while Berlin described calls to shoot down Russian aircraft entering allied airspace as counterproductive. Sweden, by contrast, said it was prepared to do so.
Three MiG-31s violated Estonian airspace for twelve minutes with their transponders switched off, no flight plans, and no radio contact, without being challenged.
A rational adversary will always choose the weakest node.
Among thirty-two countries, effective deterrence is not the average of national resolve. It is the resolve of the least determined member.
The third weakness concerns decision cycles.
Moscow decides within hours through a short, noncontradictory chain of command.
The Alliance consults, classifies, attributes, deliberates, and then communicates.
Weeks pass between the act and the response, when the useful window for retaliation is measured in days.
An organization whose reaction time structurally exceeds the adversary’s operational tempo does not deter. It comments.
The fourth weakness is economic, and it is decisive over time.
The cost ratio between a cheap attacking system and the interceptor used to destroy it is unsustainable, and the timelines are badly misaligned.
Eastern Flank Watch, the counter-drone system for the eastern flank led by Finland and Poland, begins deployment in 2026 but is not expected to reach full capability until the end of 2028. The broader spending effort runs to 2035, while incidents are occurring weekly.
No capability buildup can overcome such a temporal asymmetry, and the adversary knows it because the adversary sets the pace.
The fifth weakness is dependence on enabling capabilities, and it explains political caution better than any other factor.
What Washington is withdrawing is not infantry. It is precisely what makes action possible: intelligence, electronic warfare, aerial refueling, strategic transport, suppression of enemy air defenses, and command capabilities.
The US Air Force said in June that it would reduce its number of fighter aircraft in Europe by one-third, while the Supreme Allied Commander himself referred to an unhealthy codependence of allied forces on American forces.
A European government contemplating punitive action therefore discovers that it cannot conduct that action alone and concludes that it should refrain.
Capability dependence mechanically produces political caution, and Moscow can observe that in real time.
The sixth weakness is posture, and it governs all the others.
From antitank barriers installed in August by Polish and German units under the Eastern Shield program to anti-drone walls, the entirety of the European effort is an effort at protection.
But a shield imposes no cost.
It tells the adversary how much it will cost to get through, never how much it will cost to have tried.
That asymmetry is exactly the Russian bet.
It leaves the bill with the defender, which is the very economic objective of a strategy of harassment.
A defense that does nothing but protect itself organizes its own exhaustion, and Europe’s current strategy is nothing more than a delayed defeat financed on credit.
That assessment should not, however, lead us to overestimate the adversary, because Russia is not operating from a position of absolute strength.
Its war economy absorbs a large and growing share of the federal budget. Attrition in Ukraine remains massive, and demographics are working against it.
The use of disposable agents recruited outside the intelligence services in fact reflects constraint rather than abundance.
A service with ample resources does not outsource operations to task-paid operatives who can be identified through their DNA within three weeks.
The hybrid campaign is also the method of a power that cannot afford something more costly, which strengthens the argument rather than weakening it.
Real pressure—maritime, financial, and aerial—would hit a system already under strain.
One final point illustrates the vacuum.
The latest public opinion surveys, including YouGov polling conducted in seven countries in late June 2026 and a Reagan Institute survey released before the Ankara summit, show that Atlantic solidarity remains strong in principle.
Between 65 and 73 percent of Americans say they would be prepared to defend an ally under attack, and 73 percent believe continued US membership in the Alliance serves their own security.
But all these surveys ask about an identifiable attack—the precise scenario Moscow is trying to avoid.
No one has asked an ally, whether a public or a government, what it would do after an attributed but unclaimed act of sabotage.
Western publics have been polled about the war of 2030 and never about the war of 2026.
Finally, there remains one ultimate weakness that is neither legal nor capability-based, but political.
This silence is not an oversight. It is a choice.
It is less costly in the short term than announcing a position that would later have to be enforced.
Deterrence formulated after the fact is no longer deterrence. It is commentary.
And thirty-two nations commenting on an ongoing aggression do not constitute a military alliance. They constitute an observatory.
Tsiporah Fried — It is true that while we are getting much better at detecting hybrid attacks, we are still poor at deterring them.
However, I do not think the weakness lies solely in NATO’s conventional deterrence.
The problem is deeper. Our deterrence architecture was designed primarily to prevent identifiable aggression, while the present confrontation depends precisely on the difficulty of identifying the moment at which a succession of actions becomes aggression.
Classical deterrence works relatively well when the adversary knows that a given action will trigger a given response.
But how do you deter an attack that is deliberately ambiguous in attribution, limited in its effects, spread across several countries, distributed among state actors and proxies, and conducted simultaneously in several domains?
NATO has already substantially evolved its doctrine in this area.
As I said earlier, it recognizes that a sufficiently serious cyberattack or hybrid attack could, under certain circumstances, lead to the invocation of Article 5.
But that does not fully solve the problem.
Consider attribution.
Europeans and their allies have become much better at identifying and publicly naming those responsible for cyber or clandestine operations.
That is essential.
But attributing an attack is not the same as deterring its perpetrator.
If the consequence of an act of sabotage or a cyberattack remains essentially a political condemnation, a few diplomatic expulsions, or another round of sanctions, we must ask whether we are really changing the adversary’s cost-benefit calculation.
There is a second weakness, perhaps an even more important one: our institutional fragmentation.
A single Russian campaign can simultaneously involve intelligence services, police, the judiciary, cybersecurity bodies, critical-infrastructure protection, the European Union, NATO, and the armed forces.
Each of these actors can perform its mission perfectly well and yet collectively produce a strategically insufficient response.
Russia can have an integrated strategy against democracies whose response remains organized in silos.
That is a major asymmetry.
We probably therefore need to begin thinking not only in terms of incident thresholds but also campaign thresholds.
A coordinated accumulation of hostile actions should be assessed collectively, even when each individual action remains below the threshold for a major response.
In other words: a campaign should be assessed as a campaign, not simply as the sum of its incidents.
Atlantico — To escape the boiling-frog trap, what kind of response would actually break the dynamic of Russian provocation? Should that response come from NATO itself, from individual states, or from a combined effort?
Frédéric Dulin — Nothing will change as long as Russian action remains cost-free.
Breaking the dynamic therefore requires three things simultaneously: reestablish the link between an act and its consequence, make that link predictable and public, and accept that retaliation may take place outside the realm of direct military confrontation, where Europe possesses genuine superiority.
There are five levers, in this order.
The first lever is declaratory and costs nothing.
Publish a schedule of consequences tied to categories of acts: sabotage of critical infrastructure, attacks on an allied logistics site, attacks on an undersea cable, confirmed interference in a national election, with measures agreed in advance and triggered without a new round of deliberation.
The value of retaliation lies in its prior announcement, not in its severity.
The adversary currently exploits the contrast between the unpredictability of its own initiatives and the certainty of our inertia.
Reversing that relationship is the first genuine act of firmness.
It costs nothing in equipment or budgetary resources, but it requires a political decision.
The indispensable corollary is that Article 4 must cease to be merely a forum for consultation and become the mechanism through which predetermined retaliatory measures are implemented.
The maritime lever is certainly the most effective in the short term.
The shadow fleet finances the war by circumventing the oil price cap and simultaneously provides logistical support for undersea sabotage operations.
It offers a legally sound, economically painful, nonmilitary pressure point.
Port-state controls, systematic verification of insurance and flag validity, inspections at the entrances to straits, detentions for documentary or environmental reasons, and seizures where the legal conditions are met.
Germany has just begun moving in this direction by announcing increased pressure on the fleet.
If generalized to the Danish straits, the English Channel, and the Mediterranean, and applied consistently rather than intermittently, the measure would impose a measurable daily cost that Moscow would immediately feel, without the slightest risk of its being classified as an act of war.
The financial lever is the one Europe must use most carefully.
The €210 billion in Russian sovereign assets frozen on the continent, including €185 billion held by Euroclear, remain untouched in principal. Since 2024, only the extraordinary income they generate has been used to repay G7 loans.
But this income costs Moscow nothing, since Russia lost access to the assets when they were frozen in 2022.
A reparations loan drawing on the principal would have a material effect unmatched since the beginning of the hybrid campaign.
However, it faces an obstacle that cannot simply be dismissed as a procedural inconvenience.
Sovereign immunity protects central-bank assets, and a power that bases its entire position on international law cannot cast that law aside whenever it becomes inconvenient.
Doing so would validate the very argument Europe opposes: that the West writes rules from which it exempts itself.
It would also drive reserve holders away from the euro over the long term, at a time when multilateral institutions are already widely contested.
Belgium’s resistance therefore reflects prudence, not timidity.
But this leads to the real question.
If Europe declines to use this lever out of respect for the law, it must be all the more willing to use the tools the law clearly provides: vessel controls, individual sanctions against those commissioning hostile operations, and policing of its own airspace.
Yet those are precisely the measures Europe is failing to use.
A power that renounces a legally contested instrument while neglecting legally uncontested ones is not protecting the law. It is protecting its comfort.
The lever of unconditional sovereignty over allied airspace belongs directly to the Alliance.
Publicly announce that any unidentified aircraft entering allied airspace will be destroyed, with uniform rules of engagement from the High North to the Black Sea and tactical-level delegation of the decision.
The argument about escalation risk must be reversed.
That risk is exactly what Moscow exploits, and escalation dominance remains with Moscow for as long as Europeans deliberate when they should act.
An aircraft shot down in sovereign airspace with its transponder switched off and no flight plan is an entirely lawful act of air policing, not an escalation.
The Turkish precedent of 2015 should be recalled for what it demonstrated: an immediate and acknowledged kinetic response did not produce a war, but it did produce lasting Russian caution.
Russia retreats when it encounters firmness and advances where it encounters weakness, without a known exception since 2008.
The final lever is clandestine, and it can be exercised neither by NATO nor by the European Union.
An organization of thirty-two members operating by consensus does not conduct discreet operations; it publicizes them.
Offensive cyber operations, operations targeting the logistical and energy infrastructure supporting Russia’s war effort, active counter-interference in the Russian-language information space—all of these belong to states and to narrow coalitions built around real capabilities, involving perhaps four or five countries at most.
Symmetry of impunity remains the only language Moscow understands without translation.
None of these five levers requires firing the first shot.
All of them, however, require accepting confrontation, and that is what Europe has refused to do since 2014.
We should also be precise about what must absolutely be avoided.
A mirror-image response, action for action, gives the adversary control over the tempo and allows it to choose the terrain.
A declared red line without the means to enforce it destroys more credibility than saying nothing.
And a summit-based approach—responding to a continuous campaign with a one-off declaration—produces exactly the opposite of the desired effect because it signals that the response is tied to an event rather than a standing rule.
The division of responsibilities then becomes straightforward.
NATO should provide the shield, multidomain integration, and common rules of engagement—the one area in which the Alliance is both legitimate and capable.
States and narrow coalitions should impose costs because that requires speed, discretion, and an acceptance of risk.
The European Union should wield economic and regulatory power: frozen assets, maritime law, sanctions aimed at those who commission operations rather than the operatives who carry them out, and protection of electoral processes ahead of the 2026 and 2027 elections, the real target of the current campaign.
A combined effort, therefore, but not a merged one.
Waiting for the Alliance to do everything means accepting that it will do only what thirty-two capitals are simultaneously willing to approve—which is very little, very late.
Three or four determined states would produce more deterrent effect than twelve months of unanimous communiqués.
Vladimir Putin will stop on the day continuing costs him more than stopping.
That day will not arrive on its own.
It will be decided in Brussels, Berlin, and Paris, and every month of delay raises the price.
The choice is no longer between firmness and peace.
It is between firmness chosen now and war imposed later.
Tsiporah Fried — We must first avoid a false choice: either we do nothing or we respond militarily.
It is precisely the existence of that binary choice that favors below-threshold strategies.
The adversary knows that we will hesitate to escalate militarily in response to a limited incident.
An effective strategy therefore needs to give us more options, not fewer.
That means gradually moving away from an essentially reactive approach toward a genuine logic of deterrence or retaliation—in the Anglo-American sense—one that is integrated and graduated.
First, we need to aggregate incidents.
If several acts of sabotage, cyberattacks, influence operations, or violations are linked to the same campaign, our response must take into account the cumulative effect of that campaign.
Second, the adversary must know that this accumulation will produce a predictable increase in the costs it will have to bear.
We need a genuine ladder of responses: stronger resilience and counter-interference capabilities; public attribution and exposure of networks and proxies; economic and judicial sanctions; expulsions; restrictions targeting the structures involved; cyber countermeasures where possible and legally appropriate; strengthening of our military posture; and, at the highest level, a collective Alliance response.
The goal is not necessarily to militarize every incident.
The more credible intermediate options we possess, the less trapped we are by the choice between inaction and military escalation.
Ideally, we should be able to combine resilience, denial, and punishment.
The adversary must know that repeated hostile action will ultimately generate a cost greater than the benefit it seeks.
The remaining question is: who should act?
I think it would be dangerous to choose among states, the European Union, and NATO, precisely because they possess complementary instruments.
States must retain the ability to respond immediately. Their infrastructure, networks, companies, or territory are generally the first things affected.
The European Union has considerable instruments in the economic, financial, regulatory, digital, and legal fields. It is also essential to the protection of critical infrastructure and to economic resilience.
And NATO must provide what the other two levels cannot provide on their own: the framework of collective deterrence and, behind it, the military credibility of the Alliance.
The real challenge is therefore to make sure these three levels do not produce three parallel responses, but one strategic effect on the adversary’s calculations.
That is probably where the next evolution in Western doctrine needs to occur.
For several years, we have done a great deal of work on resilience.
We now need to devote more effort to deterrence—in the Anglo-American sense, which gives a broad role to conventional deterrence.
The objective should be not only to prevent a major military attack, but also to convince the adversary that the accumulation of hostile actions below the threshold will itself eventually produce costs greater than the benefits it seeks.
That is why I would ultimately turn the boiling-frog question around.
The temperature is not rising only because Russia is gradually increasing the level of its actions.
It is also rising because we have a tendency to normalize each one of them.
Cyberattack after cyberattack, sabotage after sabotage, incursion after incursion, we treat separately what may belong to one and the same strategic campaign.
The real question may therefore no longer be: how far can Russia go without triggering Article 5?
It is instead: how long can we continue treating as separate incidents what, for our adversary, constitutes a single confrontation?
Europe must recognize that it is not facing a succession of hybrid incidents.
It is facing a strategy of asymmetric confrontation whose nature it still does not fully understand.