On 2 March 2026, at the French Navy's Île-Longue submarine base in Brittany, President Emmanuel Macron delivered a speech that did four things in sequence. He announced the first quantitative increase in France's nuclear arsenal since 1992. He announced that France would stop disclosing the size of its total stockpile, restoring strategic ambiguity that the long-standing doctrine of "strict sufficiency" had explicitly rejected. He significantly modified that doctrine, retaining the formal framework of strict sufficiency while widening the operational scope it permits. And he committed France to "forward deterrence," a posture permitting French nuclear-armed Rafales to be temporarily deployed to the territory of named European allies for exercises and signalling. The same day, in a joint declaration with German Chancellor Friedrich Merz, France and Germany announced the establishment of a Franco-German nuclear steering group, with Germany committing to conventional participation in French nuclear exercises and joint visits to strategic sites.

Eight countries committed publicly at the speech to participate in forward deterrence: Germany, Britain, Poland, the Netherlands, Belgium, Greece, Sweden and Denmark. The list overlaps materially with the existing footprint of US B61 nuclear sharing. The official position, set out in the joint Macron-Merz declaration the same day, is that the new arrangement adds to NATO's nuclear sharing rather than substitutes for it. The operational logic of the two arrangements pulls in different directions if the United States becomes substantively unreliable, but the formal architecture as of 2 March 2026 is parallel rather than replacement.

This is the largest European defence-policy shift since the Cold War. The Anglo-American press has covered it as one of several Macron speeches that may or may not amount to anything. That coverage is wrong. Forward deterrence is not French rhetoric. It is the operational answer to the structural question that European defence has been avoiding for a decade: who provides extended deterrence to non-nuclear EU members if the United States no longer will? On 2 March 2026 France gave the first concrete answer.

What "strict sufficiency" was, and what is replacing it

French nuclear doctrine since de Gaulle has rested on a deliberately narrow proposition. France maintains the smallest nuclear arsenal compatible with credible deterrence against any aggressor. That arsenal exists to defend France, including the country's "vital interests" understood broadly, with the use decision retained absolutely by the French president. The doctrine is sovereign, parsimonious, and politically defensible inside France because it has been carefully kept out of the European multilateral debate.

The cost of strict sufficiency was that it left non-nuclear European members entirely dependent on the United States for extended deterrence. The American B61-12 warheads stationed in Germany, Italy, Belgium, the Netherlands and (politically frozen) Turkey, plus the implicit US strategic umbrella, was the structure that allowed German foreign-policy thinking to remain intact for sixty years. That structure is no longer reliable. The Trump administration has not made explicit threats to withdraw nuclear sharing, but the implicit message of every Trump-era statement on European defence has been that Europe should expect to provide its own.

Forward deterrence replaces strict sufficiency with a structure that is recognisably French in its sovereignty but radically European in its scope. The use decision remains in Paris. The warheads remain French. The aircraft delivering them remain French. What changes is that those aircraft can be temporarily based in allied territory, exercised alongside allied conventional forces, and used as visible deterrent signalling at the moment a crisis demands it. The strategic effect is that a Russian planner contemplating, say, a probing operation in Estonia must now factor in nuclear-capable Rafales potentially based in Poland, with the political authority for their use sitting in the Élysée rather than in Washington.

This is not NATO nuclear sharing. NATO sharing involves dual-key arrangements where allied aircraft deliver American warheads. Forward deterrence is exclusively French. The aircraft, the weapons, the decision are all French. Allies provide basing and exercise infrastructure. The arrangement is closer in legal structure to the British nuclear arrangement with the United States than to NATO sharing, but with France as the sovereign and Europe as the protected.

The eight-country list

The participating allies are Germany, the United Kingdom, Poland, the Netherlands, Belgium, Greece, Sweden and Denmark. The list is interesting on its own terms. Germany is the central case: the country with the most acute extended-deterrence problem, the largest economy among the participating non-nuclear states, and the political weight to make Franco-German nuclear coordination consequential. Poland is the country most exposed to the Russian land threat and the loudest political voice for European nuclear extension. The Netherlands and Belgium host existing US B61 sharing and have evidently begun preparing for the post-American architecture. Greece sits at the Aegean and Eastern Mediterranean choke points. Sweden and Denmark anchor the Nordic-Baltic frontier with Finland and the Baltic states.

The United Kingdom's inclusion is the diplomatic surprise. Britain operates its own nuclear arsenal and has not historically been considered a candidate for participation in another power's forward deterrence. The signal seems to be that the British-French nuclear cooperation framework has expanded beyond technical and exercise coordination into something closer to alignment of doctrine. The 2010 Lancaster House Treaties on nuclear cooperation and the Trinity House framework with Germany now sit inside a wider architecture that did not exist a year ago.

The countries deliberately not on the list are also significant. Italy is not, despite hosting US B61 sharing. Spain is not. Romania, the Czech Republic, the Baltic states are not. The list is the bloc's first nuclear core, eight countries explicitly under French extended deterrence, with the rest of the membership able to apply or to be added later.

Merz's same-day response

The most consequential reaction was the German Chancellor's. On the evening of 2 March, Friedrich Merz announced two things: a Franco-German nuclear steering group, and German conventional support for French nuclear exercises. The first is a formal coordination mechanism, the second is operational integration in everything except the use decision.

Merz's German political coalition has loosened the constitutional debt brake to fund defence and now has agreed to participate in a French nuclear umbrella. Both moves break post-war German taboos. Both will outlast Merz. The question of whether a successor government can or would unwind them is a real one, but the structural point is that the moves were made publicly, formally, and in coordination with the partner whose extended deterrence Germany is now relying on. Reversing requires positive action by a future government rather than passive acquiescence to an inheritance. That is a different politics.

The Italian, Polish and Dutch responses have been more cautious in tone but equivalent in substance. The eight-country list is real. The forward deterrence architecture is a treaty-track structure even though it is not a single treaty.

Why the Anglo-American press has missed this

Two reasons, both structural.

The first is that the British and American press understands European defence through the lens of NATO. The frame has been reliable for so long that anything which does not fit the frame becomes harder to see, let alone report. Forward deterrence is not a NATO development. It is not announced by NATO, not coordinated through NATO, and not described in NATO's vocabulary. The British defence press in particular treats anything that is not Article 5 or NATO-led as either an inferior shadow of NATO or as a French project to be regarded with suspicion. Forward deterrence does not fit either frame, so the coverage has been thin.

The second reason is that the Anglo-American press is, by editorial culture, suspicious of French strategic announcements. Macron's speeches on European autonomy have been treated as repeated rhetoric since 2017. The default editorial assumption has been that announcements of European defence architecture are aspirational and that nothing concrete follows. That assumption was reasonable for most of Macron's earlier speeches. It is wrong now, because the operational substance has caught up.

The European-language press, particularly in Germany, France, Poland and the Nordics, has covered forward deterrence as the major event it is. Le Monde, Frankfurter Allgemeine, Rzeczpospolita and Aftenposten have all run sustained analysis. The Atlantic Council and CSIS have published serious dispatches. Chatham House has noted that this is "Gaullist policy updated for a more unstable world." The English-language tier-one press in London and New York has not caught up to that level of seriousness.

The result is that the most important defence-policy shift in the bloc in sixty years is being processed at different speeds in different countries. France, Germany and Poland are operating as if forward deterrence is settled. Britain is reporting as if it is one of several speeches. The United States is largely not reporting it.

What it actually means

For the bloc-architecture question this column has been tracking, forward deterrence does several things at once.

First, it answers the extended-deterrence question that has been blocking European defence integration for a decade. Non-nuclear members can now have a credible nuclear umbrella from a European power, sovereign of Washington, with a decision-chain that runs to Paris rather than to a White House that is no longer reliable. This is the largest single piece of the post-American defence architecture. It was missing on 1 March 2026. It is in place, even if not yet fully operational, by sundown on 2 March.

Second, it forces the United Kingdom into a decision the country has been avoiding. The British nuclear arsenal is operationally tied to the United States through the leasing of Trident D5 missile bodies. If the US becomes structurally unreliable, that arrangement is exposed. The British options reduce to three: continue US-tied at increasing political cost, sovereignise at a programme cost in the tens of billions, or share with France. The Lancaster House framework already exists. Forward deterrence is the doorway into the third option if the British political class chooses to walk through it. The choice is now visible.

Third, it changes the structure of any future European Combined Forces architecture. Previously the operating assumption was that nuclear deterrence would remain a separate national matter. Now there is a candidate central nuclear architecture, French-anchored, eight-country participating, expandable. Any treaty-level instrument the bloc signs in the next four years will write around this structure rather than around the previous absence.

Fourth, it sharply raises the political stakes for the United States. Washington has options in response: tolerate, retaliate, or pre-empt. Tolerance maintains the appearance of a transatlantic relationship, retaliation accelerates the post-American transition, pre-emption (offering a renewed nuclear sharing commitment under Trump or his successor) is the path that would weaken forward deterrence by making it appear redundant. Watching which option the Trump administration picks in the coming months will tell us how much further the architecture can move before US response forces the pace.

The catch

French sole national authority on use is retained. This is the essential feature, not a bug. Macron's framing was specific: forward deterrence permits exercises and signalling, deployment for crisis-relevant signalling, but no transfer of decision authority. The Élysée retains the call. Critics within the bloc will argue, accurately, that this means the German Chancellor or the Polish Prime Minister cannot count on French nuclear release in their hour of crisis. The counter-argument, equally accurate, is that no extended deterrence promise is ever fully credible until tested, and that the operational presence of nuclear-capable French aircraft on allied soil is a test that does not have to be made in extremis.

Whether forward deterrence is actually deterrent in the technical sense will depend on Russian calculation. The Russian planning establishment is not naïve. They will read the eight-country list, the operational arrangements that follow, and the speed at which Franco-German conventional support to French nuclear exercises is delivered. The answer they reach in 2027 or 2028 about whether forward deterrence binds will set the pace of every other bloc decision.

Closing

What happened at Île-Longue on 2 March 2026 is the largest single piece of the post-American European defence architecture being put in place. It was announced from a submarine base in Brittany. Eight countries signed up the same day. The German Chancellor formalised participation by sundown. The architecture exists.

The Anglo-American press is reporting this as a Macron speech among several. Anyone serious about European security should be reporting it as the largest doctrinal shift in European defence since the Cold War, because that is what it is.

The sentence that has been missing from the English-language coverage is the simplest one: forward deterrence is real, it is operational, and it is the central architectural piece that was needed for a European defence bloc to function without Washington. The political conversation that follows is which country joins the eight, when the operational arrangements settle, and what the British nuclear arsenal looks like in 2030.

That conversation is happening already. It just is not happening in English.

This piece reflects independent analysis. The author writes in a personal capacity.