
In the north, NATO is rehearsing the deployment of forces to the island of Gotland, in the Baltic Sea off the coast of Sweden. Near the Polish-Lithuanian border, Belarus is conducting military exercises. Warsaw is warning of possible Russian strikes that could be passed off as an “accident”. In Paris, the president is convening political leaders and intelligence chiefs in response to stepped-up Russian activity. Switzerland — a state whose security policy has for decades been built around neutrality — is officially reporting a marked deterioration in its security environment. European governments are preparing ever more seriously for scenarios that, until recently, seemed unlikely.
On Gotland, Sweden and Germany are taking part in Exercise Silver Wotan in September. The island lies in the middle of the Baltic Sea, between the Swedish coast and the Baltic states, close to the sea and air routes that link the region. Its position gives it particular importance: from Gotland, it is possible to monitor much of the Baltic Sea and to control the approaches to its eastern seaboard.
For the Baltic states, this is not a matter of abstract geography. In a crisis, their defence will depend on how quickly allies can bring in reinforcements, equipment and ammunition. Some of those reinforcements will have to travel by sea and air; any attempt to restrict these routes would therefore immediately alter the military balance in the region.
Gotland’s importance thus lies not only in the forces deployed there. The island is one of the points on which NATO’s ability to reinforce its positions in the Baltic rapidly depends, should ordinary tension suddenly turn into a crisis.
Points of vulnerability on the map
In contemporary European security, distance is increasingly measured in time rather than in kilometres. What matters is not only where the Alliance’s forces are, but also how long it takes them to get to where their presence can still change the course of events.
A few hundred kilometres to the south-east, the same problem emerges in a far narrower form — quite literally, on the map. The Suwałki Gap, a strip of some one hundred kilometres between Poland and Lithuania, links the Baltic states to the rest of NATO territory while running between Belarus and the Russian exclave of Kaliningrad. In September, Belarusian forces rehearsed scenarios there involving mechanised units, drones and counter-sabotage operations. Minsk presents these exercises as defensive.
Yet in European military planning, geography is rarely just a matter of geography. The corridor matters not because it would by itself constitute a future front line, but because it provides the land link between Poland and the Baltic states. In a crisis, a few dozen kilometres can suddenly count for more than hundreds of pages of strategy documents: they determine whether troops and munitions can be moved, populations evacuated and communications between allies kept open. The corridor, however, is a marshy region, difficult for mechanised troops to take.
Gotland and the Suwałki Gap thus feature in the same military calculation. One opens up the possibility of action in the Baltic Sea; the other shows just how vulnerable the land link to the Alliance’s eastern flank can be. But the map ceases to be enough once the threat takes the form not of an armoured column, but of an isolated strike whose origin has yet to be established. That is precisely the problem Poland now faces.
Europe must think about its security, even far from the front line
In September, Donald Tusk said that the Polish services were considering the possibility of Russian attacks using drones or missiles that could be passed off as accidental. In such a scenario, the first strike against a state is not merely physical. It also hits its capacity to understand immediately what has just happened. A missile or a drone can be shot down. It is far harder, within a matter of hours, to establish where the attack came from, what its intention was, who was responsible and the point at which the response ceases to be a political gesture and becomes part of the conflict. For a conventional war, this sequence is unusual. For European security, it is increasingly becoming the norm.
In Paris, this shift is already viewed not merely as a matter concerning the eastern flank, but as a question of the resilience of the state as a whole. In September, Emmanuel Macron brought together the leaders of political parties and the heads of the intelligence and security services against a backdrop of stepped-up Russian activity. France’s agenda depends less and less on the geography of the front. Critical infrastructure, cyberspace, drones, foreign interference and sabotage operations are drawing a different map of threats — one on which distance from Ukraine is no longer a reliable indicator of distance from the war.
Even the incident in Leipzig, which the German authorities have linked to Russian activity, is significant from this perspective. It shows how far from the front line the space can extend in which a European state must think of its own security as part of a wider confrontation.
Swiss neutrality in the face of the threat
Switzerland is responding to the same shift from a different historical shore. For decades, its security policy was built around neutrality, territorial defence and keeping its distance from military blocs. But neutrality is no invisible shield against cyberattacks, espionage, sabotage or influence operations. Switzerland’s new security strategy notes a deterioration in the international environment. Bern is not abandoning neutrality, but it increasingly and openly acknowledges the limits of its capacity to protect itself. There is a certain historical irony in this. A state that long built its security on keeping other people’s conflicts at arm’s length must now come to terms with the fact that contemporary threats do not necessarily ask which side one is on.
States rearming
At the same time, Europe is devoting more resources to defence than at any other point in recent decades. According to the European Defence Agency, defence spending by EU member states reached around €418 billion in 2025 and is expected to rise to some €454 billion in 2026. Following the Hague summit, NATO adopted a new target: by 2035, Allies will have to spend 5% of their GDP on defence and security-related areas, including at least 3.5% on core defence requirements.
But the figures alone say little about the political resilience of the system. Defence has a peculiar property: its most important outcome is invisible. Voters see the new equipment, the bases, the contracts and the sums written into the budget. What they do not see are the wars that this spending has made it possible to avoid; the cost of deterrence therefore always carries a political asymmetry. The spending is immediate and tangible, while its benefit lies in the future and is hypothetical. With a protracted war in Ukraine, this asymmetry is becoming increasingly noticeable: governments are simultaneously funding defence, supporting Kyiv and protecting critical infrastructure, and must explain to their societies why all of this has to continue.
Robert Fico, the Slovak Prime Minister, who has said he fears a major European war and does not want his country to be drawn into a conflict through NATO’s mechanisms, illustrates not so much a rift within the Alliance as the boundary between the collective logic of deterrence and the national political cost. In London, the problem is framed differently: NATO must not only demonstrate its ability to defend its territory, it must also convince Moscow that any attempt to test the Alliance’s limits will come at a predictable cost.
The limits of military escalation
This gap between the Alliance’s strategy and national politics could prove even more significant than the defence budget figures. Collective defence exists in military headquarters, but decisions on its use ultimately pass through parliaments, governments and societies, each with its own perception of acceptable risk. It is precisely this gap that Moscow is working on.
On 11 September, in New Delhi, Vladimir Putin declared that the deployment of European troops to Ukraine would mean war with Russia. In virtually the same breath, he asserted that Russia had never threatened European states, had no intention of doing so and had no reason to enter into conflict with Europe.
These assertions may seem contradictory so long as they are regarded as mere diplomatic statements. Politically, they serve different purposes. Russia simultaneously denies posing a threat and reserves the right to decide which action by European states might give it the pretext to describe the situation as war.
For European governments, this means that the limits of escalation are not defined solely by their own strategy. They are forced to take into account how those limits will be interpreted in Moscow.
A state between war and peace
Every next step is thus caught between two risks. Inaction may convince Moscow that the line can still be pushed back. An overreaction, on the other hand, may be used as proof that it is Europe itself that is raising the stakes.
This is no longer quite the Europe in which the main defence question was whether an army could protect a border. Ports, energy grids, cables, satellite systems, data centres, railways, airspace and political institutions must now be protected — often before it is even known precisely what has been done to them.
Much of this infrastructure is privately owned, but its vulnerability has consequences for the state. A cable may belong to a business, a data centre to a company, a port to a private operator. Yet if they fail at a critical moment, the problem is no longer the company’s alone; it becomes the state’s. European defence is thus gradually turning into the defence of a system, rather than merely of a territory. Within that system, the costliest resource may be not weaponry but time: the time needed to establish where an attack came from, to consult allies and to take a political decision — and the time the adversary has left before its next move.
Gotland shows the price of speed, the Suwałki Gap the price of logistics, Poland the price of attribution, Paris how far from the front line a security crisis can now begin, and Switzerland the limits of older conceptions of neutrality. Taken together, this adds up to a reality more complex than the usual formula of “peace or war”.
Europe is not in open war with Russia. But the old conception of peace as a state in which war is simply absent describes its reality less and less. Between these two states a space has opened up in which it is possible to test reactions, impose costs, damage infrastructure, conduct intelligence operations and influence political processes, while staying far enough from the obvious threshold beyond which a response becomes automatic.
The hardest question, then, is not whether a Third World War will begin. It lies elsewhere: how long can a system keep functioning when every new incident raises the cost of inaction, yet no single episode, taken on its own, yet provides sufficient grounds for a major response?
In such a world, deterrence does not depend solely on the number of soldiers or missiles, or on the sums spent on defence. It depends on the ability of states to remain politically legible at the very moment the situation ceases to be so. And perhaps it is precisely this gap — between what has happened and what states will ultimately dare to call by its name — that will prove one of the principal tests of European security in the years ahead.
Translated from French. The French original prevails.






