By BarathVector Editorial — 2026-08-07

As of 7 August 2026, 21:00 IST.

Saudi Crown Prince Mohammed bin Salman, Turkish President Recep Tayyip Erdoğan and Pakistani Prime Minister Shehbaz Sharif signed the Makkah Joint Defence Agreement at Al-Safa Palace on 7 August. The three governments chose Islam's holiest city and the most sacred day of the Muslim week, and they chose them deliberately. The venue was itself the message, and the signatories meant it to be read as one.

Within hours the agreement had acquired a nickname. "Islamic NATO" is not an Indian coinage: it has been attached to the Saudi-led Islamic Military Counter Terrorism Coalition since 2015, and in this story it appears in a Russian International Affairs Council analysis by Fahil Abdulbasit Abdulkareem published on 27 April 2026 under exactly that title, and in the British Muslim outlet 5Pillars. It is also inaccurate, and the reason is visible on the face of the agreement. There is no council, no combined command, no standing force, no headquarters, no secretariat, no stated ratification pathway and, as of publication, no published text.

Here is the argument. The clause signed in Makkah is weak, it has already failed its only live test, and it will not be honoured as written. Saudi Arabia has not defected from the American order — it signed an American defence agreement in November and took an American nuclear framework a fortnight ago. Egypt, which fields the largest Arab army in the region, sat through four rounds of consultation and did not sign. And none of that settles the question that matters, because what governs a decision to start a war is not what your allies will do. It is what you believe they will do.

This first part sets out what was signed and what it is worth. The second takes up the harder question: why a pact this thin can still make a war more likely, and why neither Delhi nor Islamabad now holds the event that would begin one.

A note on usage, since it recurs. Where this article says "Rawalpindi" it means the Pakistan Army's general headquarters and the general staff who sit there. Where it says "Pakistan" it means the state. In defence and India policy the distinction is largely formal, because the army is the decisive actor — but it is not formal at all when the subject is 250 million people, and the words are used deliberately throughout.

The text, and the absences

The operative language is borrowed almost word for word from Article 5 of the North Atlantic Treaty: any armed attack against any one of the three states shall be regarded as an attack against them all. The agreement "provides for the enhancement of all aspects of defence cooperation" and is framed as establishing "collective deterrence against any act of aggression." That much is reported consistently across Dawn, Al Jazeera and The National.

Everything else is absence, and the absence has a precedent: the text of the Saudi–Pakistan bilateral pact of 17 September 2025 was never released either.

What Riyadh released instead was a set of denials. Rayed Krimly, the Saudi Deputy Minister for Public Diplomacy, said the agreement "does not represent an effort to establish a military axis or a sectarian bloc," is "not linked to any nuclear ambitions or arms race," "does not constitute a threat to the security of any country in the region," and — the line worth reading twice — "does not supersede or replace any existing bilateral or multilateral agreements."

Ankara was working the same seam. An Ankara-based source said the agreement "will not alter or undermine any other alliance commitments the three countries may have." A Turkish official described it as "defensive in nature," "not directed against any specific actor," and "open to other regional countries." A second Turkish official was less emollient, and this is the line Delhi should note: "The agreement will change the balance of power in the region."

A mutual-defence treaty whose own signatories announce on day one that it overrides nothing is telling you what it is for. It is not for fighting. It is a claim on three things that can be collected without anyone firing: diplomatic cover, cash, and the appearance of backing. Each of the three signatories wants a different one of them most, and that pattern — thin commitment, specific collectible — is the shape of everything that follows.

The test that already happened

This is the fact that has gone missing from the coverage of the signing, and it is decisive. The Saudi–Pakistan commitment was available to be invoked, under fire, and it was not invoked.

Ali Khamenei was killed on 28 February 2026, in strikes reported by the Washington Post, NPR, CNN and PBS as a joint American–Israeli operation. Iran retaliated across the Gulf. A missile attack on Riyadh airport was intercepted on 1 March. Drones over the Ras Tanura refinery were intercepted on 2 March; the facility was hit by a projectile on 4 March and did not reopen until 13 March. A drone struck the American embassy in Riyadh on 3 March. The SAMREF refinery was hit on 19 March and Prince Sultan Air Base on 27 March. Three Saudi civilians were killed and 23 injured over the campaign. On 2 March the IRGC closed the Strait of Hormuz to what it called unfriendly nations.

If a mutual-defence pact means anything, it means something then.

What Pakistan did was remind Tehran that the agreement existed, and then spend six weeks as a mediator. Islamabad offered itself as a venue on 23 March. Deputy Prime Minister Ishaq Dar shuttled between capitals, reportedly persuading Iran that Saudi soil would not be used against it. Pakistan and China floated a five-point plan on 31 March. Pakistan tabled a 45-day ceasefire proposal on 6 April that President Trump accepted before his own deadline. The Islamabad Talks opened on 10 April.

Only on 11 April did Pakistani aircraft reach King Abdulaziz Air Base, and Pakistani officials attached a public disclaimer to the deployment on the day, saying it was not directed at any party. The force that eventually assembled — reported at 8,000 personnel rising to 13,000, with roughly 16 aircraft, drone squadrons and a Chinese HQ-9 battery — arrived after the first ceasefire had taken hold, financed by Riyadh, with that disclaimer attached.

An analysis published by Harvard's Belfer Center before the war had assessed the 2025 bilateral pact as a political signal of solidarity rather than an unconditional war guarantee. The record of the war bears that out. Pakistan could be both mediator and Saudi ally only for as long as its deployment stayed defensive, time-bound and visibly limited; the moment the theatre turned offensive, the dual role would have collapsed. It never collapsed, because Pakistan never let the theatre shift. Under pressure, an alliance reveals its price, and this one revealed a low one.

There is a harder reason underneath, and it is the one Indian coverage keeps missing because it looks inward rather than at the treaty. Pakistan did not decline to fight Iran only because the clause was thin. It declined because it cannot fight Iran. Estimates of Pakistan's Shia population range from roughly a tenth to a fifth of the country, which makes it the second-largest Shia population in the world after Iran's own. An openly anti-Iranian war fought on Saudi instructions is not a policy any Pakistani general staff can sell at home, in a country already fighting an insurgency in Balochistan, a revolt in Pakistan-occupied Kashmir and a sectarian campaign that murdered 31 Shia worshippers at prayer in February. Mediation was not the clever option. It was the only one.

That constraint is worth holding on to, because it governs everything in the second part of this article. Pakistan presents itself to the Gulf as the Arab world's available Muslim army. What it actually delivered in the worst regional crisis in a generation was six weeks of shuttle diplomacy and a garrison that arrived after the shooting paused with a public notice attached saying it was not there to fight. The gap between the offer and the delivery is not hypocrisy. It is arithmetic — and it means the one adversary against which Pakistan's general staff can reliably mobilise its own population is the one that is neither Shia, nor Muslim, nor internal.

The pressure has not lifted. The Islamabad Memorandum that Pakistan brokered on 17 June collapsed on 8 July. President Trump declared the ceasefire over, Central Command has run hundreds of strikes across consecutive nights, the naval blockade of Iran was reinstated on 14 July and announced on 15 July, and Iran has struck Bahrain, Kuwait and Qatar while resuming attacks on shipping. At least 14 American service members are dead — a figure the Pentagon revised down from 18 on 24 July. The signature in Makkah was made in the middle of a live war, by three governments who know exactly what the clause would cost if anyone tried to cash it.

Which invites the obvious objection, and it is the strongest one available against everything in the second part of this article. If Pakistan is the party that demonstrated the guarantee was hollow, why would Rawalpindi ever over-read it?

Because Rawalpindi is not buying the guarantee. It is buying three things that survive the guarantee being worthless. It is buying diplomatic cover, of the kind Ankara supplied in May 2025 and no clause was needed to obtain. It is buying a financial floor, because a treaty makes the deposits that hold up Pakistan's IMF programme harder to withdraw. And it is buying the appearance of backing, which is a currency that spends domestically whether or not it is honoured abroad. Sushant Sareen of the Observer Research Foundation, whose reconstruction of Pakistani war calculation is the subject of part two, supplies the piece that closes the gap: the international support Rawalpindi expects in a short war with India is a belief rather than a calculation. A belief does not have to survive contact with March 2026. It only has to survive contact with a general staff reasoning about a war it expects to last five days.

The date nobody in Delhi is quoting

On 18 November 2025, two months after signing with Pakistan, Mohammed bin Salman went to the White House, signed a Strategic Defense Agreement with the United States and was conferred Major Non-NATO Ally status the same day. Washington is processing F-35 sales to Riyadh.

In February 2026 the administration notified Congress that it was pursuing a civil nuclear pact with Saudi Arabia without the non-proliferation safeguards the United States has traditionally required. The Additional Protocol, which gives IAEA inspectors access to undeclared sites and which the UAE accepted alongside its 2009 agreement with Washington, is not in the draft. On 22 July 2026 the 123 Agreement was signed by Energy Secretary Chris Wright and Prince Abdulaziz bin Salman, alongside a 30-year framework, with Westinghouse to build AP1000 reactors.

So Riyadh took an American defence agreement in November, an American nuclear framework a fortnight ago, and an Islamic collective-defence clause on 7 August. That is not a country defecting from the American order. It is a country stacking every available insurance policy after discovering, on 4 March, that the one it already held did not stop a projectile from reaching Ras Tanura.

It also means India should expect no American sympathy in objecting to any of this. American non-proliferation analysts have been blunt about the Saudi nuclear track: the Council on Foreign Relations warns it would set a dangerous new proliferation precedent, and the Bulletin of the Atomic Scientists describes it as dragging Washington into non-proliferation quicksand. The proliferation risk that Indian commentators are attributing to Makkah was authored in Washington, and it is the one Israel is actually worried about.

Three signatories, three different enemies

The pact is routinely described as uniting an oil exporter, a NATO member and a nuclear-armed state. True, and analytically useless. The revealing question is who each of them is afraid of.

No two of them agree on the enemy. No two of them share a patron. The Turkish entry is not abstract: Netanyahu has accused Erdoğan of having "massacred" his own Kurdish citizens, Erdoğan has warned that Turkey will not tolerate fantasies of a "Promised Land," and in late March and early April 2025 Israeli aircraft struck three sites in central Syria that Turkish teams had surveyed for a future airbase, cancelling a planned Turkish visit. A collective-defence clause among three states in that configuration is a mutual insurance policy against abandonment, taken out by three parties who each expect to claim rather than pay.

Their motives are correspondingly specific and cheap. Saudi Arabia is buying manpower and deterrence-by-association that does not run through Washington, and paying in cash: it has extended a $5 billion facility to Pakistan to December 2028 and financed the deployment on its own soil. Türkiye is buying leverage. A Chatham House assessment published in January argued that Ankara's search for alternatives to NATO affords not only alternatives to alliances but, crucially, leverage within them, and that such arrangements offer Türkiye a way to shore up its regional power, build its export base and develop its ballistic missile technology. The wound is seven years old: the S-400 purchase in 2019 and the F-35 expulsion that followed.

Pakistan is buying money and status. Its IMF programme — a 37-month, $7 billion facility — is conditioned on Saudi, Chinese and Emirati deposits of roughly $12 billion staying in place: $5 billion from Riyadh, $4 billion from Beijing and $3 billion from Abu Dhabi. It raised $27.2 billion in external financing in FY26 counting rollovers, but only $16.2 billion of that was fresh, against a target of $19.92 billion — short by $3.72 billion. What it gets in Makkah is a treaty that makes those deposits harder to withdraw.

The most eloquent fact about the pact's ambition is who is not in it. Egypt attended every round of the parallel four-way foreign ministers' track — Riyadh on 18 March, Islamabad on 29 March, Antalya on 17 April, and at least four rounds by July — and fields 438,500 active personnel, the largest Arab army in the region. It sat through all of it and did not sign. Its reasons are not on the record, so its absence is inference rather than statement; what the absence establishes is the ceiling. No member of that quadrilateral has extended collective defence to another and none has proposed doing so, which means the arrangement signed on 7 August is not the nucleus of a bloc. It is the largest configuration that was available, and the largest Arab army declined to be in it.

How the world is reading it

Washington has said nothing, as of publication. There has been no White House, State Department or Pentagon statement. That is silence rather than refusal, and it should be distinguished from the State Department's position on the September 2025 bilateral pact, which it expressly declined to discuss. Two instincts are pulling against each other and neither has won: alarm at an arrangement that dilutes American leverage over three partners at once and normalises Pakistani extended deterrence in the Gulf, against relief that allies are burden-sharing while American forces fight Iran with 14 dead and a blockade to sustain.

The UAE answered by signing with Delhi, seven months early. On 19 January, Sheikh Mohamed bin Zayed spent three hours in New Delhi and signed a letter of intent on defence and security cooperation, an agreement to conclude a framework for a Strategic Defence Partnership, nuclear cooperation on large reactors, a $3 billion LNG deal making India Abu Dhabi's largest customer, a target of $200 billion in trade by 2032, and a joint condemnation of terrorism using India's own formulation, "including cross-border terrorism." Foreign Secretary Vikram Misri described it as a natural evolution of the relationship. It is more than that: it gives India the benefit of Gulf alignment with none of the liability of an attack-on-one clause. That is not an accident of this one agreement. India has spent seventy years declining to sign the thing every other middle power eventually signs — it is in no mutual-defence treaty, belongs to no bloc with an automatic clause, and has never accepted an obligation to fight another state's war. It is the reason that when the Gulf burned this spring, India had nothing to discharge and nobody to answer to. The signatories in Makkah have each bought an insurance policy they hope never to claim. India declined to sell one.

The other capitals converge on a single reading, that the clause is real enough to matter politically and too thin to matter militarily. Israel declined to comment, as it did in September 2025; its exposure is nonetheless the greatest of any capital, and it is not the Makkah clause but the fact that Pakistan holds an estimated 170 warheads, has not signed the Non-Proliferation Treaty, and fields missiles that reach Israel. Iran is publicly dismissive: Ibrahim Rezaei, spokesperson for the Iranian parliament's National Security and Foreign Policy Committee, said reliance on foreign agreements "will not constitute a safety valve for Riyadh" — a hardening from September 2025, when President Pezeshkian welcomed the Saudi–Pakistan pact as the start of a comprehensive regional security system. China is silent and gaining: John Calabrese, writing for the Middle East Institute in "Silent leverage, quiet gains?", assesses that if any country stands to benefit from the pact without assuming its risks, it may be China — though analysts also judge that Beijing will not enable Saudi nuclearisation through a Pakistani intermediary, because it would alienate two fellow SCO members in India and Iran. Russia has said nothing; Abdulkareem's RIAC paper of 27 April 2026 analyses the earlier four-way track rather than the 7 August signing, which it predates by three months, and reads it as structurally weak — different threat assessments, no institutionalisation, economic weakness, political instability, and a grouping that "runs a risk of being driven by events." Moscow's comfort with all of this is an inference from its conduct, not a position it has stated. And the smaller Gulf states are hedging in every direction: Qatar holds an American executive-order security commitment and is simultaneously in talks with Pakistan, while Kuwait ratified a defence cooperation agreement with Pakistan by decree-law in 2026. Both were struck by Iran in July, because both host American bases.

What it costs India

Three consequences survive the theatre, and the ranking most Indian coverage uses is upside down.

The industrial one is serious, and it is the only one that compounds. Treaties do not transfer combat power; production lines do. Türkiye has sold Pakistan MILGEM-derived corvettes, supplied the drones that flew over Indian positions in May 2025, is reported to be exploring a drone plant inside Pakistan, and is courting Islamabad for the KAAN fifth-generation fighter programme. The claim of a $15 billion, 65-aircraft KAAN deal rests on a single Turkish journalist and deserves the scepticism Pakistan's fiscal position demands — a country $3.72 billion short on its external financing target does not casually find $15 billion. But Burcu Özçelik, a researcher at RUSI, identified the agreement's real content when she noted that it creates significant opportunities to deepen Türkiye's defence-industrial footprint through joint production, technology partnerships, procurement and greater military interoperability. A defence clause can be ignored in an afternoon. A shared assembly line cannot, and the difference between those two facts is a decade of Indian planning assumptions.

The second is friction rather than force, and it is the one place the alliance has already been honoured. In a crisis Pakistan does not need Saudi or Turkish troops. It needs what Ankara supplied in May 2025: diplomatic cover, resupply, and the closure of Turkish airspace to Indian equipment in transit. That was a political commitment kept, and it was kept without any defence clause at all — which is the sharpest available demonstration that the clause is not what generates the support.

The third is status, and Delhi should find it uncomfortable. Pakistan brokered a US–Iran ceasefire, hosted the talks, and sat between the Saudi crown prince and the Turkish president on 7 August. India practised what it called calibrated neutrality through the war, and the region read it as absence. The Chabahar sanctions waiver lapsed on 26 April with no replacement.

What does not survive is the encirclement arithmetic. On SIPRI's calendar-2025 figures, published in April 2026, India spends $92.1 billion on defence, fifth in the world. Pakistan spends $11.9 billion and Türkiye $30 billion, and neither is transferring its budget to Rawalpindi. China, which actually supplies Pakistan's inventory, signed nothing.

And India's gravest West Asian emergency of 2026 was never a treaty. It was a strait. When the IRGC closed Hormuz on 2 March, India — which imports about 88 per cent of its crude, with roughly 45 per cent of it, half its LNG and 90 per cent of its LPG transiting that channel, according to the Atlantic Council — watched its weighted average crude price go from $69 in March to over $114 in April. The IEA called it the largest supply disruption in the history of the oil market. India holds about 74 days of strategic and commercial stocks against the IEA's 90-day benchmark, and its dedicated strategic reserve alone covers 9 to 10 days of net imports. That is what a bad month in West Asia costs. Three signatures on an unpublished document do not move a barrel.

Riyadh, meanwhile, is hedging toward Delhi as openly as it hedges toward Islamabad: a $100 billion investment framework signed during the Prime Minister's Jeddah visit in April 2025, some $43 billion in annual trade, 2.65 million Indians in the Kingdom among 9.7 million across the Gulf who send home close to 40 per cent of India's bank remittances, a Strategic Partnership Council co-chaired by the Indian prime minister and the crown prince with a ministerial defence pillar, and joint exercises at Sada Tanseeq and Al Mohed Al Hindi.

The Ministry of External Affairs has pitched its response accordingly. Asked about the trilateral on 7 August, spokesperson Randhir Jaiswal said only that "this is also a development that we are closely following, and we shall keep you updated." The fuller Indian formula was set out in September 2025, in response to the bilateral pact: that the arrangement formalises something long-standing, that India would study its implications, and that India expects its strategic partnership with Riyadh to be pursued "keeping in mind mutual interests and sensitivities." No outrage, no summoning of envoys, no boycott campaign. Acknowledge, decline to escalate, and remind the wealthiest signatory what it stands to lose.

Why a thin pact can still be a dangerous one

Everything above argues the pact is thin. None of it argues the pact is safe.

Take the fairest case for the other reading first. Abdulaziz Sager of the Gulf Research Center gave the most careful characterisation available on the day: that three of the Muslim world's most influential states convening at a moment of heightened uncertainty demonstrates a growing willingness among regional and middle powers to coordinate on security. His forward assessment is entirely conditional — "if institutionalised and translated into concrete cooperation," the framework could strengthen the three countries' collective diplomatic and defence weight and contribute to a more regionally driven security architecture. That conditional is doing all the work, and on the record of the last five months there is no institutionalisation to point at. But Sager is right that the willingness is real, and a reading that treats the pact as pure theatre has to explain why three governments spent political capital on it.

Here is the explanation this article offers, and it produces an unwelcome correction to the alarmed Indian reading and the reassuring one alike.

Sushant Sareen's assessment for the Observer Research Foundation, published on 14 July 2026 under the title Pakistan on the Ropes: Why a Cornered Military May Gamble on Short War With India, reconstructs the Pakistani calculation as follows. Rawalpindi assumes a shooting match would not last more than a few days. It expects international support from the Arab world, Saudi Arabia, Qatar, Türkiye and the West. It expects President Trump to intervene on its behalf. It expects a brief conflict closed by American mediation.

Read that list against the pact. It does not have to be honoured to do damage. It only has to make that list look more plausible to the people holding it.

Which means the pact probably lowers the chance that a war, once started, runs long — there are more capitals with money at risk in a South Asian conflict than at any point in twenty years, and Riyadh, Ankara, Beijing and Abu Dhabi would all want it over by day five. And it probably raises the chance that one starts, because a war that looks cheaper is a war that is easier to choose. Both effects are real. Only one of them is being discussed.

Part two examines the second effect: the condition of the army that would make that choice, the religious frame it has adopted and why that frame is failing, the narrative campaign it has come to believe, and the specific sequence by which a terror attack on Indian soil now escalates faster than it did in May 2025.