22 September, 2026
|
15 minutes to read
On 7 August 2026, in a palace in Mecca, three heads of government signed a document stating that an armed attack against any one of them would be regarded as an attack against all. Turkish Foreign Minister Hakan Fidan told Reuters the next day that the mutual defense element was technically the same as NATO’s Article 5. Türkiye, Saudi Arabia and Pakistan share no borders with one another. They sit arrayed around Iran, which Türkiye and Pakistan border by land and Saudi Arabia meets across the Persian Gulf, and which had been under American and Israeli attack for five months when the signatures were affixed. Three states with no common frontier had bound themselves to a collective defense clause built around a shared adversary rather than a shared perimeter.
The division of labor in this new alliance runs on complementary functions, with money from Riyadh, defense industry from Ankara, and combat manpower and an unstated nuclear status from Islamabad. Pakistan also becomes the first state outside the Non-Proliferation Treaty to extend a mutual defense clause to partners holding no nuclear weapons, in a field where formal extended deterrence has been almost exclusively an American instrument.
What matters here is less whether the clause would ever be activated than what these three governments judged they had to offer one another, that they could no longer get from anyone else.
There are reasons to question what this really means. The text has never been published, and neither has the Saudi-Pakistani agreement of September 2025, whose formula it reproduces almost word for word, suggesting a document was inherited rather than drafted by the three parties. There is no joint command, no standing force, and no trigger. What exists is a ministerial committee, political and military committees, a permanent secretariat in Saudi Arabia, and operational procedures still to be worked out. The agreement covers the state-level strike of the kind Israel delivered against Doha in September 2025, through a mechanism that commits nobody to anything automatically, and says nothing about the constant sub-threshold pressure its members actually live with. Every one of these objections may have merit, but the agreement still tells us something important, because what matters here is less whether the clause would ever be activated than what these three governments judged they had to offer one another, that they could no longer get from anyone else.
Alliances have always been priced in three currencies. Territory, for the depth, basing, and denial it provides. Capability, for reaching a fight. And political identity, which decides who gets admitted to the club. NATO admits European states that can uphold the principles of the Washington Treaty and contribute to North Atlantic security. Its stated foundations are democracy, individual liberty, and the rule of law. The Collective Security Treaty Organization (CSTO) drew its members from the former Soviet Union, placing greater weight on geographic proximity and dependence.
Mecca revalues all three. Territory seems to be largely irrelevant, since no two signatories share a frontier. Capability broadens from forces to functions, with capital, industrial base, manpower, and nuclear ambiguity spread across the three countries. The political criterion is the most interesting. The text names no enemy, and officials from all three governments described the agreement as purely defensive and open to any brotherly nation that wishes to join. That phrase does the work in this arrangement that democracy does in the Washington Treaty, setting eligibility by shared faith rather than by shared political order or governance model. The summit was convened at the invitation of the Custodian of the Two Holy Mosques, and the Pakistani prime minister performed the Umrah hours before signing.
The Mecca agreement raises three questions: what alliances have historically bought, what changed to make this one thinkable, and what any of it means for states that cannot secure themselves alone.
Set against NATO and the CSTO, the Mecca agreement raises three questions: what alliances have historically bought, what changed to make this one thinkable, and what any of it means for states that cannot secure themselves alone. The Black Sea is where those questions are least theoretical. Two of its littoral states sit outside every collective defense arrangement while absorbing continuous pressure from the largest, and the region holds assets that no alliance framework in existence knows how to price.
Alliances were designed against that sequence, on the reasonable assumption that preventing the conquest prevented what the conquest was for.
Historically, the value of these three currencies followed a clear strategic logic. For most of the modern era, alliances were built around a particular understanding of how a state loses its independence. To exercise political control over another country, you had to defeat it, occupy it, install a government that would do what you wanted, and build the institutions through which the arrangement would run. Conquest was the entry cost, and everything that mattered politically came afterward. Alliances were designed against that sequence, on the reasonable assumption that preventing the conquest prevented what the conquest was for.
This is a claim about which route was primary rather than about which routes existed. Control without invasion was always available, and Tehran in 1953, Guatemala in 1954, and Santiago in 1973 remain the standard illustrations. Soviet doctrine theorized and used subversion at length, and the Finnish arrangement institutionalized one version of it. The non-military route lacked reliability. It moved slowly, cost a great deal, depended heavily on conditions inside the target, and could be undone. Conquest was faster and more certain, so the defensive architecture was built against conquest.
Territory earned its price in that world during that time because it did three things at once. It supplied the depth that allowed a defense to trade space for time. It provided the bases and infrastructure through which reinforcement arrived. And it denied the adversary the same ground. Contiguity mattered because proximity made assistance physically feasible and because neighbors facing the same adversary perceived the threat in the same way, which is harder to fake than a signature.
Capability earned its price for reasons that need less explanation. Alliances are promises to fight, and a member that cannot reach a fight or contribute to one is a liability whose defense others must underwrite. The concern runs through every enlargement debate NATO has held.
Political identity is the currency that gets least attention and does the most work. The Washington Treaty does not admit every European state that contributes to North Atlantic security. It admits states in a position to advance the treaty’s principles, which the preamble names as democracy, individual liberty, and the rule of law. The CSTO projected a different identity, drawing its members from a single dissolved union and binding them through dependence on Moscow rather than through any shared political order. Both arrangements restricted entry on political grounds. They differed in which politics they required.
The assumption underlying the old price tags has been failing in public for four years, and 2026 made that failure hard to dispute.
Consider first what overwhelming force now buys. The United States held complete conventional dominance in Afghanistan for two decades and departed, leaving the government it had gone in to remove. Russia has held conventional superiority over Ukraine since February 2022 and occupies roughly a fifth of the country, having produced in the process a Ukrainian polity more firmly Western in orientation than the one it attacked and two additional NATO members on its own border. The United States and Israel opened the most intense air campaign of the decade against Iran on 28 February 2026 and killed its supreme leader within days. Six months on, the political settlement that campaign was meant to produce has not arrived, and its most visible consequence has been a public quarrel inside the alliance about who helped and who did not.
That quarrel is itself instructive, and the record is contested enough to require care. Trump stated in March that most allies had declined to join, called the alliance a paper tiger and its members cowards, and said in April that NATO had not been there when the United States needed it. NATO Secretary General Mark Rutte countered on 24 June with a figure of up to five thousand American military flights launched from NATO territory during the six-week campaign, including from Germany, the United Kingdom, Italy and Romania. Romania approved base access on 11 March through its Supreme Council of National Defense and parliament, hosted roughly five hundred additional American troops with tankers and surveillance aircraft, and closed Bucharest airport to commercial traffic so American tankers could launch. Spain refused the use of Rota and Morón and closed its airspace on the ground that the operation fell outside the UN Charter; Italy withheld a Sicilian base because its bilateral agreement required parliamentary approval that could not be obtained in time; and Poland declined only to relocate Patriot batteries, on the stated ground that they protect Polish skies and the eastern flank. What divided the alliance was legal opinion, parliamentary procedure and, above all, political judgment. Military capability was not the binding constraint in a single case.
Now consider what control costs when nobody fires. Hungarian voters ended sixteen years of Fidesz government on 12 April 2026, returning a constitutional majority to Péter Magyar’s Tisza party. Those sixteen years had carried a NATO and EU member a considerable distance from the political order it joined, through electoral rules, control of the judiciary, media dominance, and the use of state resources. No foreign soldier was involved at any point, and neither alliance possessed an instrument that addressed the process. The reversal came the same way, from an election, with the alliance again playing no part.
Georgia is the case where both halves can be observed against the same adversary pursuing the same objective. Russia won militarily in August 2008 in five days, but produced a country more firmly Western in orientation than the one it invaded. The campaign conducted from 2012 onwards, without a shot fired, delivered what the invasion could not – a subservient government, which effectively satisfies all of the Kremlin’s objectives in the region
Political control of another state no longer requires taking its territory, which means the instrument alliances were built to deter is no longer the primary route to the outcome alliances exist to prevent. Territory retains value, but its value is being realized in a different way.
None of this means geography stopped mattering. Something more specific happened. Political control of another state no longer requires taking its territory, which means the instrument alliances were built to deter is no longer the primary route to the outcome alliances exist to prevent. Territory retains value, but its value is being realized in a different way.
Two arrangements in a single month, in regions that share nothing, illustrate the point. On 18 September 2026, Trump announced that the United States had agreed with Denmark and Greenland to give Washington permanent control over security and all other needs in Greenland, with no expiry and no cost, and barring American adversaries from bases, military presence, or sensitive investments without express approval. Nothing has been signed. Both governments confirmed a deal and expect to sign it in New York during the General Assembly, but they describe it differently. Danish Prime Minister Mette Frederiksen stated that it does not give the United States sovereignty, that it recognizes the Kingdom’s sovereignty and territorial integrity and the Greenlandic people’s right to self-determination, and that it strengthens common security in the Arctic and North Atlantic and is therefore good for NATO and Europe. On the record, a gap remains between an American claim of permanent control and a Danish insistence on retained sovereignty, with a signature still to come.
The dynamic is worth more than the text. Denmark spent this year under annexation pressure severe enough that its military drew up an operations order dated 13 January contemplating the demolition of Greenlandic runways with explosives to stop American aircraft landing. It resolved that pressure by conceding security control over the territory to a fellow member of its own alliance through a bilateral arrangement negotiated entirely outside the alliance.
The Mecca arrangement defines a set of states whose defense is a matter for that set alone. The Greenland deal, as Washington describes it, reserves to the United States a permanent veto over who may hold a base, a presence, or a sensitive investment on the island. Neither involves Russia, and neither was negotiated through an institution. Both accept that security is settled by powers within defined zones rather than by rules applying equally to everyone, which this journal has previously identified as Moscow’s central objective, and which now appears on its own in regions where Moscow has no part to play. That earlier analysis warned in April 2025 that American rhetoric on Greenland and Canada was normalizing the use of national security to justify territorial ambition. Eighteen months later, the normalization will have a document attached to it.
Spheres of influence are returning to the agenda across several regions at once, and the states most exposed are those that sit inside a contested zone where no primacy has been established.
If the currencies have been revalued, the question for a state that cannot defend itself alone is what it now holds that a guarantor needs and cannot obtain elsewhere. Several answers have become available that did not exist in the old pricing, and they are concentrated in the Black Sea.
Corridors are valued for what crosses them rather than for the depth they provide. Georgia holds the only land route between the Caspian and the Black Sea that avoids both Russia and Iran, which is a scarce asset in a year when both powers are subject to sanctions and both are unreliable as transit. Operational knowledge has become tradable because the character of war changed in four years, and Ukraine holds Europe’s only body of war-tested drone and counter-drone practice. Exposure to continuous sub-threshold attack generates defensive experience that the rest of the alliance lacks, and more than seventy documented violations of NATO and Moldovan airspace, including a cruise missile that came down near Lublin on 29 July, have made frontline states the only ones accumulating it. Energy and data transit carry their own value, and Romania’s Black Sea gas and Georgia’s pipeline and fiber routes are both instances. Public legitimacy belongs on the list as well, since a guarantor’s willingness now turns partly on whether a partner’s population will sustain the alignment under pressure.
Armenia has replaced its suppliers and gained monitoring, funding, and consultation mechanisms. But it has not acquired a guarantor.
Armenia is an interesting case, too. Yerevan froze its CSTO participation in February 2024 and stopped contributing financially in May of that year. It passed legislation initiating EU accession in January 2025, agreed to a Strategic Partnership Commission with Washington the same month, diversified arms procurement toward France and India, signed a framework allowing participation in EU crisis management operations, and held first consultations with Brussels on security and defense focused on hybrid threats and disinformation. On 7 June 2026, a government pursuing that course won a contested parliamentary election against a Russian-backed alternative, taking 49.81% and sixty-four of one hundred and five seats on the highest turnout for a scheduled election since 2018, while all three opposition parties entering parliament campaigned on repairing ties with Moscow. Armenia has replaced its suppliers and gained monitoring, funding, and consultation mechanisms. But it has not acquired a guarantor, and no process with a defined endpoint has opened, raising the question of whether this contribution, while necessary, is actually sufficient.
The Mecca agreement formalized something that was already load-bearing. Pakistan and Saudi Arabia have run a military relationship for six decades, structured by a 1982 protocol that stationed more than ten thousand Pakistani troops in the Kingdom, and by mid-2026 Pakistan had again deployed several thousand troops along with fighter aircraft, drones and an air defense system. Türkiye arrived with an umbrella already extended to Qatar, Syria and Libya. The document did not create any of that. It recorded a set of relationships that had become dense enough to be worth recording, which is how alliances have generally come about. They are written over existing substrate, and the currencies examined here are simply the forms that substrate takes.
A corridor is only worth something if the state it crosses can guarantee what passes through, an alignment is only worth something if the government means it, and a constitutional obligation is only worth something if the bodies bound by Article 78 intend to execute it.
That is the part of the Mecca lesson most likely to be missed in Tbilisi and in Kyiv. Neither country is standing outside a door, and neither is in a queue whose order somebody else controls. The operative question is what each is accumulating now that a guarantor would later find worth formalizing. Ukraine has been accumulating it under fire since 2022, in a body of operational knowledge that no European military possesses and every European military will need. Georgia held a comparable set of assets and has spent three of them, because a corridor is only worth something if the state it crosses can guarantee what passes through, an alignment is only worth something if the government means it, and a constitutional obligation is only worth something if the bodies bound by Article 78 intend to execute it.
In a system reorganizing around zones of recognized primacy, the Black Sea is the one contested region where no power has established one, which is precisely why the pressure on it is continuous. Whether it becomes someone’s sphere by default depends on what its states build before the question is settled elsewhere.