Pakistan brachte Iran zum Ja. Trump hat es in 5 Tagen verbrannt.

Key Takeaways

• A deal existed for five days. On April 7, Pakistan’s Prime Minister Shehbaz Sharif and Field Marshal Asim Munir talked Trump out of a threatened civilization-ending bombing run and into a two-week ceasefire. Trump accepted “the general framework” of Iran’s 10-point proposal as a basis for negotiations. Pakistan brokered it. The deal lasted until April 13.
• Trump trashed his own ceasefire during a public feud with the Pope. On Sunday, April 12, Pope Leo XIV called the Iran threats “truly unacceptable” and diagnosed a “delusion of omnipotence.” Trump replied by calling the Pope “weak,” posting an AI-generated image of himself as Jesus, and the next morning ordering a naval blockade of every Iranian port. Whether the feud caused the reversal or merely overlapped with it, the ceasefire died inside the news cycle.
• The CIA has now given Trump 90 days. A leaked May 7 assessment delivered to the White House concluded Iran can survive the blockade for “at least three to four months” with 70% of its missile stockpile intact. The 90-to-120-day window expires before the November midterms and inside what mediators describe as still-active backchannels routed through Islamabad and Muscat.
• The problem is no longer the broker. Pakistan is still trying. Oman is still trying. The problem is finding terms that survive Trump’s signature. The 1981 Algiers Accords held because Algeria built parallel, separately-binding obligations that Iran and the United States could each commit to independently, with Bank of England custody of the assets, an architecture that did not require the two sides to trust each other. No such architecture exists for 2026. Trump’s April 13 reversal compounds the difficulty for any broker trying to assemble one inside the CIA’s 90-day window.

The Tuesday That Worked

On the morning of Tuesday, April 7, the President of the United States told ninety-three million Iranians that “a whole civilization will die tonight” unless Tehran surrendered control of the Strait of Hormuz by midnight Tuesday. The threat covered every bridge and every power plant in Iran. It was the most extreme presidential rhetoric against a foreign nation in modern American history. The US did this to Iraq in 1991: twenty-eight power plants destroyed, power output dropped to 4% of pre-war levels, an estimated one hundred thousand civilians dead from collapsed water and sanitation systems.

By 12:42 PM Eastern the same Tuesday, Trump posted a different statement to Truth Social. The threat was off. A ceasefire was on. The post named two people:

“Based on conversations with Prime Minister Shehbaz Sharif and Field Marshal Asim Munir, of Pakistan, and wherein they requested that I hold off the destructive force being sent tonight to Iran, and subject to the Islamic Republic of Iran agreeing to the COMPLETE, IMMEDIATE, and SAFE OPENING of the Strait of Hormuz, I agree to suspend the bombing and attack of Iran for a period of two weeks. This will be a double sided CEASEFIRE!”

Iran’s Foreign Minister Abbas Araghchi confirmed the deal hours later in language that named the architecture. The United States, he said, had submitted a fifteen-point proposal. Iran had submitted a ten-point counter. The President of the United States had publicly accepted “the general framework of Iran’s 10-point proposal as a basis for negotiations.” Trump called the same document “a workable basis on which to negotiate” and claimed “almost all of the various points of past contention have been agreed to.”

Strip away the bluster. The structure underneath was a textbook mediation. Pakistan supplied the back-channel between Tehran and Washington. The two sides converged on Iran’s framing. Trump publicly signed off. The Brent contract crashed from $109.77 to $93.00 inside one session, a 15.3% single-day decline, as the oil tape priced in the de-escalation. The deal was real. The market believed it. The only question was whether the United States President believed it.

What It Took the Pope to Find Out

The answer came in five days.

On April 12, Pope Leo XIV (Robert Prevost, the Chicago-raised Augustinian who had become the 267th Bishop of Rome the year before) was asked about Trump’s “civilization will die” threats. He called them “truly unacceptable” and a symptom of the “delusion of omnipotence” driving the war. He went further when asked about Trump personally: “I have no fear.”

Trump’s response was the rupture. He called the Pope “weak,” “terrible,” and “very liberal.” Then, late on the night of April 12, he posted an AI-generated image of himself in flowing white robes, hand outstretched on a hospital patient’s bed, golden light radiating behind him, a nurse and a soldier flanking the scene. The image was deleted within hours. Screenshots spread across every platform on earth.

By 10 a.m. the next morning, April 13, the US Navy began enforcing a blockade of all Iranian ports in the Strait of Hormuz. The ceasefire was less than six days old. The Pakistani mediation, the Iranian 10-point proposal, the “general framework” Trump had publicly accepted, all of it was reframed inside one news cycle as a blockade his Navy was now enforcing.

The public record does not show what triggered the reversal. The administration did not cite cabinet pressure or new intelligence about Iranian non-compliance with the strait reopening; no such reporting has surfaced. What is public is the sequence: a Sunday-night Pope feud and AI Jesus image, followed by a Monday-morning blockade order. Whether the feud was the trigger, the cover, or the coincidental backdrop, the deal that took a foreign army chief one Tuesday afternoon to broker did not survive the weekend.

Mediators Are Still Working. The Problem Is the Signature.

A month later, the mediator inventory is not actually empty.

Pakistan’s foreign minister told NPR on May 6 he was “hopeful a U.S.-Iran deal can happen soon,” in language consistent with quiet backchannel work routed through Islamabad and Muscat. Oman has spent the past fifteen years hosting the most consequential US-Iran back-channels in modern diplomatic history, including the eight rounds of secret Muscat talks between 2012 and 2013 that produced the JCPOA framework. Sultan Qaboos, and his successor Sultan Haitham, built that channel by treating discretion as the product.

What is missing is not the broker. It is the architecture that lets the broker’s work translate into a deal that holds.

Compare the 2026 picture to the 1981 endgame of the original hostage crisis. The Algiers Accords (signed January 19, 1981, freeing fifty-two American hostages) did not work because Algeria persuaded both sides to trust each other. Algerian Foreign Minister Mohammed Benyahia and his team proposed a structurally different agreement model: each country undertook its obligations independently of the other, as separately binding declarations, rather than as a bilateral treaty. The Bank of England held the frozen Iranian assets in escrow (the account was in the name of the Central Bank of Algeria). The Iran-United States Claims Tribunal at The Hague was created to litigate the financial disputes for decades afterward, and it is still operating in 2026. Iran did not have to trust the United States to honor its word. The United States did not have to trust Iran. They trusted the architecture, which was custodied by neutral third parties with reputational and legal capacity to hold the line.

The April 7 Pakistani-brokered ceasefire had no such architecture. It was a public commitment by two heads of state, articulated in Truth Social posts and a single statement from Iran’s Supreme National Security Council. When Trump reversed it on April 13, there were no escrow accounts to release, no third-party tribunals to invoke, no Bank of England-equivalent custody to embarrass him. The only consequence was that Iran, Pakistan, and the watching world learned that an American presidential commitment to this deal had a shelf life of one weekend.

That is the verification problem. And it is the binding constraint, not mediator availability.

What Carter Had That Trump Does Not

The four institutions that powered the 1981 Algiers Accords have either eroded, been bombed, been insulted, or been replaced by belligerents in 2026.

1981 Architecture	2026 Equivalent	Status
Algeria as neutral Sunni mediator	Pakistan / Oman backchannels	Active but stripped of public weight after April 13 reversal
Swiss “Protecting Power” of US interests in Tehran (since 1980)	Switzerland (still de jure)	Neutrality corroded since 2022 EU sanctions alignment
Bank of England custody of Iranian frozen assets	No equivalent custodian	None exists; Treasury OFAC has been politicized as a daily policy lever
Iran-United States Claims Tribunal at The Hague	No equivalent forum	The Hague is treated as an adversary by the current administration
Qatar / UAE Gulf-state diplomatic cover	UAE and Qatar are active war participants	UAE hit at Fujairah on May 4; Qatar’s LNG terminals were drone-struck on March 2
Pope John Paul II (sent Vatican envoy to Tehran during the 1979-81 crisis)	Pope Leo XIV	Personally insulted by Trump April 12; “AI Jesus” image circulated the same night

Read down that column and the structural problem becomes legible. Algeria worked in 1981 not because Algerian diplomats were unusually persuasive, but because the four supporting layers (Swiss neutrality, British financial custody, an international tribunal, and a Gulf diplomatic cover) were all functioning. In 2026, those four layers are either non-functional or have been actively converted into combatants in the war the mediator is being asked to end.

The Pew Research poll out of late April makes the political reality stark from the American side. Sixty-two percent of Americans disapprove of Trump’s handling of the Iran military action; 45% strongly disapprove. Even among Republicans, 32% disapprove. A president with that approval profile cannot credibly commit to a multi-year tribunal process; he might not be in office when the proceedings start.

The Boring Hypothesis: Maybe Deals Just Take Time

The case against the mediator-vacuum thesis is straightforward: deals take time. The JCPOA took two years of secret Muscat talks before the public negotiations began. Treating 90 days as a deal window may itself be the misreading. Maybe the May 7 CIA memo just describes a normal diplomatic clock that, like all diplomatic clocks, runs longer than political calendars want.

The boring hypothesis has a real point. But the 1981 Algiers timeline cuts the other way. The decisive Algerian mediation produced a signed accord in roughly ten weeks of intensive work between November 1980 and January 19, 1981. The reason ten weeks was enough is the same reason ten weeks worked: the architecture existed. Swiss, British, and Hague-based institutions had standing legal capacity. The mediators did not have to invent the verification layer; they only had to fit the specific deal terms into it. 2026 has neither the institutions nor a counterparty whose signature is worth the verification.

That is not a deal problem. It is a system problem. And it does not get solved in 90 days from a standing start.

The Only Broker That Can Force a Deal Has a Different Price Tag

The one player with structural enforcement capacity over Tehran, the ability to make a deal stick by threatening to remove what Iran depends on, is Beijing. The Iran-China 25-year strategic cooperation agreement, signed March 27, 2021, covers trade, security cooperation, banking, finance, insurance, and energy investment, with China designated as a steady importer of Iranian oil. Through the 2026 war, Chinese refineries have continued absorbing Iranian crude via shadow-fleet relabeling, with cargoes arriving on Chinese soil routinely relabeled as Malaysian, Omani, or Emirati to obscure origin.

Beijing’s hold is therefore not rhetorical. It is contractual. A credible threat to slow or restructure the 25-year cooperation agreement would change Iranian behavior in ways that no Pakistani back-channel can match.

The price of that mediation, however, looks structurally hard for Trump to pay. Beijing’s terms would likely include some relief on the US export control regime the current administration treats as existential, the same regime the site has analyzed in the chip-whiplash piece. Trump’s electoral coalition may absorb four-dollar gasoline through November. Trading away the chip standoff to settle a different war is a much harder sell. The mediators that are politically acceptable in Washington have limited hold over Tehran. The mediator with real hold is the one whose price tag looks hardest for this administration to pay. The “ceasefire on life support” sits inside that vise.

What 90 Days Without a Mediator Looks Like

The leaked CIA memo’s 90-to-120-day window is not a negotiating clock. It is a deterioration clock. Iran can absorb that level of economic punishment without political collapse, with most of its missile stockpile and 75% of its mobile launchers still operational, for the duration. The CSIS analysis on the April ceasefire munitions status reached the same conclusion through a different door: the air campaign degraded specific systems but did not strip Iran of the capacity to retaliate.

What happens inside those 90 days without a working signature mechanism is, on the recent pattern, fairly predictable. Pakistani and Omani backchannels are likely to continue. More “hopeful” statements will probably leak from foreign ministries in Islamabad and Muscat. The May 4 reaction in Brent (a 6% intraday move to $114 on the Fujairah strikes, paring back near $110) is the template for what episodic Iranian retaliation does to the oil tape. Absent a deal, US gas prices look set to stay above $4, and Republican congressional polling looks set to continue its slide. Every few weeks a new “framework” can be expected to leak; none of it produces a signed instrument while the architecture for a signed instrument does not exist.

The next signed deal, when it eventually comes, will be brokered by Beijing on Beijing’s terms, or it will be brokered by Trump’s successor under a different administration with a different theory of what an American signature is for. Tehran negotiated in January 1981 only after Carter was definitionally a lame duck and Reagan was inaugurated. The pattern is what the May 7 CIA memo means by “Iran’s leadership has gotten more radical, determined and increasingly confident they can outlast U.S. political will.” The regime is reading the calendar it read in 1981. The calendar reads November 3.

The Five-Day Verdict

The case for despair about the 2026 ceasefire is not that mediation failed in some abstract sense. It is that mediation worked, exactly once, on the Tuesday afternoon of April 7, and was destroyed within a single business week by the same president it had rescued from his own threat that very morning. Pakistan does not need to be re-recruited. Oman does not need to be re-discovered. The frozen-asset escrow mechanism does not need to be re-invented. They all need a counterparty whose signature is worth more than the next news cycle.

Pope Leo XIV’s phrase was not chosen as diplomacy. It was chosen as diagnosis. A delusion of omnipotence does not negotiate; it gives press conferences. The 90-day CIA window is the institutional verdict on whether the patient comes out of the delusion before the political clock runs. The 1981 precedent says the regime in Tehran has waited out an American president with this exact muscle memory before, and the leaked memo says it intends to do so again. The infrastructure for an off-ramp is filed in a Bank of England vault somewhere, on paperwork from 1981, while the current architecture for a signed ceasefire (three Truth Social posts and a Telegram thread between Islamabad and Tehran) did not survive a Sunday night and the Monday morning that followed it.

The deal Pakistan got Iran to was real. The president who burned it has 90 days. The mediator who can rebuild the architecture in that window does not exist.