How the 2015 JCPOA was built and broke apart, and what the 2026 Islamabad MOU is trying to rebuild on its ruins — an integrated analysis of where an agreement is signed, the mechanism meant to enforce it, and the actors who will not sign it.
In June 2026, when President Trump remarked from the Oval Office that a US–Iran agreement "could be signed in Europe," one thing flashed to mind: the first interim agreement of the Iran nuclear talks, back in 2013, had been signed in Geneva. And so it came to pass. [1] On 15 June, Iranian Deputy Foreign Minister Gharibabadi announced that the MOU had been finalized and would be formally signed on Friday, 19 June, in Geneva, Switzerland.
The venue is no accident. In diplomacy, the place of signature is a declaration of an agreement's lineage. The Geneva of 2013 was where a five-step blueprint began — interim deal (JPOA) → defer the core disputes → main negotiation → final agreement (JCPOA) → UN Security Council endorsement. The Geneva of 2026 reaches for that same blueprint — utterly different in content, but identical in the skeleton of its procedure.
The purpose of this document is not a mere recap of events. It is to trace a single structural repetition across thirteen years: the same city, the same procedural grammar, the same Security Council terminus. Yet between 2015 and 2026, three things changed decisively — the configuration of the veto was inverted, the nuclear facilities were destroyed by bombing so that the very baseline for verification vanished, and the resolution meant to enforce any deal must now be built atop a legal schism within the P5. These three shifts are what determine why "signing is not the end but the beginning."
Nuclear diplomacy is heavy on jargon but simple in structure. Iran holds nuclear capability; in exchange, the West imposes sanctions (economic blockade). A negotiation is the trade of "how much Iran rolls back its nuclear program / how much the West lifts sanctions." The real question is who enforces that trade. The heart of this document is exactly that — where the enforcement mechanism sits, and why it keeps failing.
To understand this MOU, one must first see how its prototype — the JCPOA — was assembled.
The JCPOA's turning point is often dated to Rouhani's 2013 election, but the real opening came earlier. Oman's Sultan Qaboos was at its center. A mediator who had cultivated decades of good relations with the US, Saudi Arabia (Sunni), and Iran (Shia) alike, Qaboos earned the Obama administration's trust by brokering the 2011 release of three American hikers detained after straying across the Iraqi border into Iran. [2] He leveraged that trust to appoint himself the back channel for a US–Iran rapprochement.
The first secret contact came earlier than commonly assumed. In early July 2012 — eleven months before Rouhani took power — Jake Sullivan, then director of policy planning at Hillary Clinton's State Department, flew secretly to Muscat to meet Iranian officials. [2a] Going back further, mid-level US officials had already been meeting Iranian counterparts in Muscat since 2011. The true turning point, then, was not Rouhani's election but Obama's re-election in November 2012 — with the political room a second term afforded, Obama activated a secret bilateral channel to the Ahmadinejad government in earnest at the outset of his second term.
The card the US played was decisive. Washington signaled a shift in position: it could tolerate a limited domestic enrichment program conditioned on intrusive verification. This was a retreat from the "zero enrichment" principle the US had held until then, and it became the decisive inducement for Iran to come to the table. Qaboos captured this progress in a letter to Ahmadinejad stating that the US representative had recognized Iran's right to enrich uranium, and sent a copy of the same letter to Obama. [2] Even so, skepticism ran inside Tehran — a distrust that "Washington cannot be trusted to keep its commitments" was already present then, and was confirmed as fact by the 2018 US withdrawal (§4) and the 2025 snapback (§5).
The June 2013 election and August inauguration of the centrist Rouhani accelerated the flow. Right after Rouhani took office, Deputy Secretary of State William Burns and Sullivan held further secret meetings that began shaping the framework of an initial nuclear deal, and at the September UN General Assembly, Obama and Rouhani spoke by phone — the first direct contact between the two countries' leaders in 34 years. [2a] This shift in the US position, together with Rouhani's election, set the stalled P5+1 (US·UK·France·Russia·China + Germany) process moving again.
The first fruit emerged in Geneva. [2] Iran agreed to limit "any further advances" of its activities — specifically, to eliminate its stockpile of 20% enriched uranium, halt enrichment to that level, and freeze its 3.5% stockpile. In return came modest sanctions relief. The crucial point is the sequence: Iran acted first, the IAEA verified, and only then did money flow. This "performance first, money later" structure is precisely what gets inverted thirteen years later in the 2026 MOU (see §7).
After the JPOA the parties met countless times and repeatedly extended their deadline. [3] The talks, led by US Secretary of State Kerry and Iranian Foreign Minister Zarif, survived a breakdown scare in November 2014 and pushed the deadline to 30 June 2015. [4] A framework was reached in April 2015, and the final deal was struck in Vienna on 14 July 2015.
These twenty months were long because five hard problems were tangled together at once. Knowing how each was resolved makes clear what is missing from the 2026 MOU (especially missiles and verification).
① Enrichment capacity — the single hardest issue. The West initially wanted "zero enrichment," but Iran fixed its right to enrich as a matter of sovereignty. [4a] The settlement was sealed in numbers — centrifuges cut to about one-third, enrichment capped at 3.67%, and the low-enriched-uranium stockpile limited to 300kg. This combination aimed to stretch the "breakout time" (the time needed to weaponize) to about a year. (For reference, by 2025 Iran had reached 60% and 6,000+kg — figures that show the scale of the deal's collapse.)
② The Arak heavy-water reactor — closing the plutonium path. Blocking uranium enrichment alone was not enough. The Arak heavy-water reactor was a "second road" that could extract enough weapons-grade plutonium for one or two weapons a year from its spent fuel. [4b] The West wanted it shut; Iran refused; the resolution was to redesign the core to block plutonium production. (This provision worked so well that even after the US withdrew in 2018, it issued sanctions waivers so the UK and China could keep modifying Arak.)
③ PMD — the political landmine of "possible military dimensions." The most politically sensitive issue was the "Possible Military Dimensions" of Iran's past nuclear work. [4a] The IAEA had to clarify whether Iran had pursued weaponization research in the past, and Iran resisted fiercely, citing military pride and secrecy. Resolving PMD was the key stumbling block in the final stretch, and was ultimately sealed by a compromise — "focus on future verification rather than fully excavating the past" — a point hardline critics cite as a "hole" to this day.
④ The sequence and speed of sanctions relief. Iran wanted "full relief upon signing"; the West wanted "phased relief after verified performance." [4c] The settlement was the latter — separating nuclear, missile, terrorism, and human-rights sanctions, lifting only the nuclear-related ones and leaving the rest, with relief taking effect only after Implementation Day, once the IAEA verified Iran's prior steps. The 2013 JPOA principle of "performance first, money later" carried through to the final deal (how this sequence is revived and distorted in the 2026 MOU is §7).
⑤ Sunsets — the deal's time-limited design. Putting expiry dates on the core constraints was the last issue and the biggest controversy. [4d] Centrifuge limits were set to expire after 10 years, the 3.67% enrichment cap after 15 (to which the 5-year arms-embargo and 8-year missile sunsets of §3 are added). What Trump cited as grounds for the 2018 withdrawal was precisely these sunsets — that "once the clock runs out, Iran legally approaches the bomb." The very device that made the deal possible (sunsets) also became the pretext for breaking it.
The JCPOA comprised a main text and five technical annexes (nuclear, sanctions, civil cooperation, the Joint Commission, and the implementation plan). [5] The core bargain was clean — Iranian nuclear constraints and transparency ↔ removal of US, UN, and EU nonproliferation sanctions. Six days later, on 20 July, the Security Council unanimously adopted Resolution 2231, endorsing the JCPOA. [5]
An important legal detail bears noting. The JCPOA text itself was not a legally binding treaty but a bundle of "political commitments." [44] Its binding force came from the Security Council resolution, 2231, that endorsed it. This dual structure — "a political agreement (JCPOA) plus the Security Council resolution (2231) underpinning it" — is decisive. The US could "withdraw" from the JCPOA in 2018 precisely because it was not a treaty, and 2231's snapback matters so much because the deal's true legal spine was the resolution.
One more thing: a separate US domestic ratification process was entangled. Obama treated the JCPOA not as a "treaty" requiring Senate ratification but as an executive agreement, and Congress held only review and disapproval powers under the Iran Nuclear Agreement Review Act (INARA). [44] That is, because the deal was from the start an executive product lacking full congressional approval, it carried the structural fragility of being reversible by a single executive order once power changed hands. Then, via Adoption Day (18 October 2015), sanctions relief took effect on Implementation Day in January 2016, once the IAEA confirmed Iran's prior steps.
The legal pillar holding up the JCPOA was Resolution 2231, and the invention inside it was snapback. Because the fate of the 2026 MOU will also be decided on a variant of this device, its mechanics are worth getting exactly right.
Ordinarily, the Security Council works on the logic that "to impose sanctions, you must pass a resolution." So if any one of the five permanent members (P5) casts a veto, the reimposition of sanctions is blocked. If Russia or China shields Iran, the West can do nothing. Resolution 2231 inverted precisely this logic. [6]
Normally, "to impose sanctions you need yes-votes," and a single veto can block them. Snapback flipped this around — now "to prevent sanctions you need yes-votes." If someone files a complaint that Iran broke its promise, a 30-day clock starts, and unless a resolution to "keep sanctions relief in place" passes, the old sanctions automatically come back. The complaining state need only veto that "keep relief" resolution. In other words, the design makes it impossible to use a veto to block the restoration of sanctions.
Concretely: when a JCPOA participant notifies the Council of Iran's "significant non-performance," a 30-day countdown begins. [7] If within those 30 days the Council fails to pass a resolution to "continue sanctions relief," six prior sanctions resolutions from 2006–2010 (1696, 1737, 1747, 1803, 1835, 1929) are automatically restored. The complaining state simply vetoes that "continue relief" resolution, or lets the vote collapse.
This mechanism is not an abstract reading but is spelled out in the resolution itself. For the sake of reliability, here is the core language set directly against its meaning.
"…within 30 days of receiving a notification by a JCPOA participant State of an issue that the JCPOA participant State believes constitutes significant non-performance of commitments under the JCPOA, [the Security Council] shall vote on a draft resolution to continue in effect the terminations… of the provisions of [the previous resolutions]."
Reading: Two things are fixed. ① The right to pull the trigger belongs to no one at large but only to a "JCPOA participant State" (the "standing" dispute of §4 and §6 originates here). ② Upon notification, the Council "shall vote" within 30 days on a resolution to continue relief — i.e., unless the Council affirmatively votes to "continue," the termination is undone. The phrase "continue in effect the terminations" is the crux. [45]
"If that motion… does not pass within 30 days of notification, then all UN sanctions lifted under the agreement are restored. This reinstatement is automatic and cannot be stopped by a veto from a permanent member of the UNSC, but only by the passage of an alternative resolution."
Reading: The restoration is automatic, cannot be stopped by a permanent member's veto, and can be blocked only by "the passage of an alternative resolution." This is the legal substance of "inverting the veto." [46]
Breaking the two sides of Figure 2 into procedural units makes the difference even clearer.
How many JCPOA participants are there? Paragraph 10 of 2231 names them — China, France, Germany, Russia, the UK, the US (P5+1), plus Iran and the EU. [45] Any one of these except Iran, acting alone, can initiate snapback. Crucially, even a country that is not a permanent Council member, like Germany, has standing (in 2025 the E3 actually triggered it — §5). Conversely, when the US — which withdrew in 2018 — tried to trigger it in 2020, it was rebuffed with "you are no longer a participant," precisely because of this paragraph-10 definition of "participant" (§4).
How does the normal Council procedure work? Ordinarily, imposing or restoring sanctions requires passing a new resolution with 9+ affirmative votes and no permanent-member (P5) veto. So a single veto from Russia or China blocks the sanctions. In peacetime this structure protects Iran.
How does snapback invert this arithmetic? Step by step: [7][46]
① Notification. One participant notifies the Council in writing that Iran is in "significant non-performance." At that moment the 30-day clock starts. (In 2025 the E3 first passed through the JCPOA's Dispute Resolution Mechanism (DRM) beforehand — §6.)
② The 10-day rule. If no Council member submits a "continue relief" draft resolution within 10 days of notification, the Council president must table that draft and put it to a vote. This is a device to prevent the dodge of letting time run out by avoiding a vote.
③ The 30-day vote. For that "continue relief" resolution to pass within 30 days, it needs — just like any ordinary resolution — 9 affirmative votes and no P5 veto. But the complaining state simply vetoes that very resolution. Then "continue relief" fails.
④ Automatic restoration. If "continue relief" fails to pass within 30 days, the six prior resolutions are automatically restored. The only way to block restoration is to pass an "alternative resolution" — which is likewise blocked by the complaining state's veto.
In the normal procedure, a veto is a shield that blocks "a resolution to impose sanctions." In snapback, a veto becomes a spear that blocks "a resolution to continue the lifting of sanctions" — the same veto works in reverse. So the complaining state, by merely holding its veto and staying silent for 30 days, brings the six back to life. What can block it is not the P5, but only a majority willing to actively pass a "continue relief" resolution.
This device was the core argument that sold the deal to the US Congress in 2015 — "we can revive UN sanctions at any time, without Russia's or China's consent." [6]
Snapback can be likened to a trial where the default verdict is "guilty." The moment a complaint is filed, unless the jury actively declares "not guilty" within 30 days, a guilty verdict (restored sanctions) is confirmed automatically. And the complaining prosecutor (the notifying state) holds a veto over any "not guilty" verdict — so the prosecutor need only stay silent for the acquittal to become impossible. In normal times a veto is "a shield that blocks sanctions"; inside snapback, it becomes "a spear that blocks the lifting of sanctions — i.e., that revives them."
Snapback was not had for free. The price Russia exacted for accepting this powerful device was the sunset clauses written into the resolution. A sunset is a time-limited clause: "this restriction expires automatically after X years." Resolution 2231 contained two such time-limited restrictions on Iran.
The crux is the asymmetry of time. Russia's carrots — the arms embargo (2020) and missile limits (2023) — still lay in the future back in 2015. But as of 2026 those sunsets have all expired, and Resolution 2231 itself lapsed in October 2025. Russia has already collected everything it was owed. Keep this symmetry in mind, and it becomes clear why a "let's plant a new snapback" bargain is so hard to strike in 2026 (see §11 and §13).
As one analyst put it, the JCPOA was "a carefully constructed machine designed to put Iran's nuclear ambitions in a transparent, verifiable box for over a decade." [8] For a few years the machine worked. The US Director of National Intelligence assessed in 2017 that the JCPOA had "enhanced the transparency of Iran's nuclear activities" and extended the time Iran would need to produce weapons-grade fissile material from a few months to about a year. [9] Defense Secretary Mattis confirmed in 2017 that Iran was complying.
In May 2018, the first Trump administration unilaterally withdrew from the JCPOA and launched a "maximum pressure" sanctions campaign. [10] Here a common misconception must be cleared up. The US did not stop sanctioning. If anything, it sanctioned harder than ever. But there were two kinds of pressure tools.
① US unilateral sanctions (OFAC) — imposed under US domestic law. They need no Security Council consent and the president can switch them on and off by executive order. Trump turned these to the maximum — "maximum pressure" aimed at driving Iran's oil exports to zero.
② UN sanctions (the ones snapback restores) — imposed by Security Council resolution, binding the whole world. Yet Trump declined to pull this snapback trigger.
Why leave the most powerful tool untouched? The usual explanations are mostly wrong. It was not "because criticizing Iran was awkward" (Trump denounced Iran daily). It was not "because of Israel" (Israel actually wanted snapback). The real reason was a logical self-contradiction.
The right to trigger snapback belongs, under 2231, only to a "JCPOA participant State." But in 2018 Trump had loudly proclaimed, "we are leaving this deal." For a state that has left the deal to wield a tool inside that deal is a contradiction — like shouting "I've quit the club!" and then insisting "but I'll still ring the emergency bell that only members may ring." By choosing the political theater of a noisy withdrawal to erase a predecessor's legacy, Trump forfeited the standing to use the most powerful weapon inside that very deal. [11]
Then in 2020 something urgent arose: the arms-embargo sunset (expiring 2020). That October, the conventional-arms embargo on Iran would lapse automatically, letting Russia and China sell arms to Iran legally. The US wanted to stop this. So in August 2020 it abruptly reversed itself, claiming "actually we are still a participant since we are named in the resolution, and therefore have standing to snap back," and attempted to trigger it.
The result was a diplomatic humiliation. Thirteen of the Council's fifteen members rejected it, on the ground that "you left in 2018, so you have no standing." [11] Even allies like Britain, France, and Germany declined to side with Washington. The US protested that it was "named in the resolution," but was ignored. [12]
About a year after the withdrawal, from 2019, Iran began rolling back its nuclear commitments in retaliation. [10] It raised enrichment levels, reaching 60% enrichment — close to weapons-grade (90%) — by 2025. [12]
The lesson made by this 2020 episode is decisive. Snapback was designed to be "automatic," but if a political dispute erupts over "who has standing to pull the trigger," that automaticity itself can be suspended. However the text reads "automatic," the premise of that automaticity — standing — remains in the realm of politics. Five years later this very logic is reused in the opposite direction — this time by Russia and China against the E3 (see §6).
There were US re-entry attempts, too. The Biden administration held several rounds of "compliance-for-compliance" talks in Vienna in 2021, but these stalled after 2022. [8] Meanwhile Iran's nuclear knowledge advanced irreversibly, and its "breakout time" shrank from a year to a matter of weeks. [8] The JCPOA had effectively entered life support.
18 October 2025 was originally to be the "Termination Day" of Resolution 2231. [13] After ten years, the remaining UN and EU nuclear-related sanctions would vanish and the snapback device would sunset. It did not play out that way.
After the June 2025 Israeli and US strikes on Iran's nuclear sites (Fordow, Natanz, Isfahan) — the so-called "12-Day War" — Iran suspended IAEA cooperation and the whereabouts of its 60% enriched stockpile became opaque. [14] The E3 (UK, France, Germany) formally triggered snapback on 28 August 2025. [12] The cited grounds were excess enriched-uranium stockpiling, restricted IAEA monitoring, and abandonment of the Additional Protocol. [14]
Within the 30-day window, Russia and China attempted a "continue relief" vote, which failed (4 in favor — Russia, China, Pakistan, Algeria), and a six-month extension of 2231 also collapsed. As a result, on 27–28 September 2025, the six prior sanctions resolutions were restored en masse. [12] The arms embargo, the enrichment/reprocessing ban, the ballistic-missile ban, asset freezes, cargo inspection — the entire 2010 architecture came back to life.
Yet Russia and China, together with Iran, declared this restoration "null and void" and refused to implement it. [12] Their grounds were twofold — ① the E3 themselves had failed to perform their JCPOA obligations (sanctions relief) and therefore lacked "participant" standing, and ② they had skipped the JCPOA's dispute-resolution procedure, a procedural defect. [12] The very logic that had denied US standing in 2020 was now applied to the E3.
And on 18 October, Resolution 2231 itself lapsed. [15] Here a decisive legal vacuum opened. The six restored resolutions originally had no sunset clauses, so on the Western reading they remain valid indefinitely. On the Russia/China reading, the restoration was void, so no UN-level sanctions on Iran exist at all. [15] Over a single Security Council resolution, the P5 stands frontally divided — a state unprecedented in UN history.
The two grounds on which Russia, China, and Iran declared the 2025 snapback "legally void" are not surface rhetoric but finely engineered legal arguments. Understanding them is the key to why the 2026 resolution is so fraught. The two arguments work as a set — ① you have no standing to complain, and ② even if you did, you skipped the procedure.
Start from the premise. The JCPOA was, in essence, an exchange of mutual obligations — Iran constrains its nuclear program, and the other side (US, E3, etc.) lifts sanctions. A bilateral contract in which both sides bear obligations simultaneously. When the US withdrew in 2018 and re-imposed sanctions, one side of that contract (the West) began failing its obligations. Every dispute flows from here.
The right to trigger snapback belongs, under 2231, only to a "JCPOA participant State." Not just any country may file a complaint. So "who is genuinely a participant" becomes the key gate to snapback's validity.
An important distinction here: the E3's "non-performance" is not an affirmative act like casting a veto, but an omission — a failure to do what they had promised to do. After the US withdrew in 2018, the E3, fearing US secondary sanctions (which punish even third-country firms that deal with the US), effectively severed economic relations with Iran. They built a workaround payment channel called INSTEX, but it barely functioned. So, the Russia/China/Iran argument runs, the E3 merely nominally maintained their JCPOA promise to "lift sanctions" while in substance isolating Iran economically. "For a party that won't keep its obligations to cherry-pick only the rights of the agreement (snapback) is a contradiction."
This logic is an exact copy of 2020. Back then, thirteen states blocked the US with "you withdrew, so you're not a participant → no standing." In 2025, Russia, China, and Iran turned that very logic on the E3 — "in 2020 you told the US it 'was not a participant'; by that logic, the E3, having failed to perform, are not participants either." The West counters that "the US declared withdrawal, but the E3 never once declared withdrawal." That is, the West treats "whether withdrawal was declared" as the test of standing, while Russia, China, and Iran treat "whether obligations were substantively performed" as the test. With no international tribunal to adjudicate, it collapsed into a political contest of strength.
Built into the JCPOA is a staged procedure called the Dispute Resolution Mechanism (DRM) (JCPOA paragraphs 36–37). [33] Its core idea is: "do not detonate the snapback nuclear bomb straight away — first pass through a dialogue stage." Even if a participant judges that another has broken its promise, it must clear this checkpoint before notifying the Security Council.
Combine the two arguments and you get a double line of defense — "no standing (①), and even with standing, wrong procedure (②)." If either is accepted, the snapback is void. And the essence of these two blades is a restatement of the lesson from §4 — even a device that the text makes automatic can be neutralized once denied at the political gates of "standing" and "procedure." This produced the P5 split of §5, and that split is the root of the resolution-drafting fight to come in §11. The reason Russia and China insist "we cannot accept language that concedes the restoration was valid" is precisely that they have argued all along that "the restoration was void for defects of standing and procedure."
After a fresh war that began in late February 2026 (triggered by US–Israeli strikes on Iran), negotiations resumed under Pakistani mediation, and on 14 June both sides officially confirmed the MOU. [16] One correction is warranted. The official mediator and political face is Pakistan (PM Sharif announced the deal and chaired the "Islamabad Talks"), but much of the operational working channel ran through Qatar — an Iranian delegation (including the central-bank governor) negotiated the unfreezing of assets in Doha, and the June 2025 ceasefire that ended the 12-Day War was a joint US–Qatar mediation. [43] In other words, "Islamabad" is the political stage and Doha the operational plumbing — a dual structure in which the leverage for the next phase (especially the funds/assets channel) may rest more with Qatar. Yet the two sides describe the same document in opposite terms. First, the US version.
The core principle the Trump administration secured is "relief for performance." [17] A senior US official put it plainly.
This is the revival of the 2013 JPOA structure from §2. In 2013 it was "performance first, money later"; this time Iran demanded "$12 billion upon signing" but failed to secure it. Vice President Vance also stressed that "the Iranians are not receiving any cash, and no funds are released merely for signing or attending a meeting." [18] The funds were ultimately weakened into a vague, back-loaded structure releasing assets "based on progress in negotiations." [19]
On the US account, the MOU commits Iran never to pursue a nuclear weapon and to negotiate the suspension of its enrichment program and the disposition of its highly enriched uranium (HEU) stockpile. [20] A White House official told Fox News it was a performance-linked deal under which Iran must dismantle its nuclear program, destroy and remove nuclear material, and stop funding proxies before receiving sanctions relief. [14]
On 14 June, Trump explicitly proclaimed a "toll-free opening of the Strait of Hormuz" and ordered the immediate lifting of the naval blockade. [21] For 60 days the strait is to open with no tolls and Iran is to clear its mines. [20] Along the line of "tolls → service-fee rebranding → control," the text ultimately settled on the US "no tolls" version. A US official stressed reciprocity — "the faster the Iranians clear the mines and let shipping resume, the faster the blockade is lifted." [20]
In short, the US version of the MOU is "a performance-linked document that extracted a chain of Iranian concessions." Markets priced it in at once, with international crude prices falling as the expectation of a lifted blockade and a normalized Hormuz quickly drained the supply-risk premium. [16]
Iran's official narrative is the opposite. Deputy Foreign Minister Gharibabadi framed the MOU as "the fruit not only of diplomacy but of Iran's military achievements during the war," declaring that "the enemy that launched the attack failed in all its malicious objectives, and the Islamic Republic achieved major victories in the war." [22] He said Iran had incorporated all its key positions into the draft and would publish the full text after signing.
The most striking phrase is this one.
This single phrase compresses everything traced in §3–§6. Having lived through the 2018 US withdrawal and the 2025 snapback, Iran has learned that "neither a Security Council resolution nor a US promise constitutes a guarantee." So it goes to the UNSC — while distrusting that very UNSC. Gharibabadi added that the armed forces "will always keep their fingers on the trigger against any enemy conspiracy." [23]
Inside Iran, however, this narrative met fierce challenge. Nabavian, deputy chair of parliament's National Security Commission, read out the original text and accused the negotiating team of "surrendering to the US." [24] He reads almost every item — Hormuz, the nuclear file, the money, the withdrawal — as "American demands presented in Iran's language," i.e., the result of Iran retreating with each successive version. His central revelation was that, in the Hormuz clause, "Iranian control" had vanished from the text and the US had inserted the word "unlimited" for passage.
Indeed, hardliners staged street rallies in Tehran and Mashhad in June, branding Foreign Minister Araghchi an "infiltrator," [25] and seven lawmakers of the "Super Revolutionaries" refused to sign a statement endorsing the negotiating team. [26] Iran's "victory" narrative, in other words, is not domestically settled.
The hardliner claims above — especially "insertion of 'unlimited' passage," "evaporation of Iranian control," and "disappearance of the $12bn upfront payment" — support a load-bearing analysis (the trigger of a US–Israel rift), yet they rest heavily on a single partisan source (Nabavian's hostile reading, Paydari-Front-aligned outlets, Iran International). Critically, even the 14-point draft published by Mehr News has not been independently verified, and neither Iranian nor US officials have publicly confirmed its contents. [19] Whether these revelations are the actual text, or hardliner spin (in Iran's own phrase, the circulation of "worst-case scenarios") meant to sink the deal, cannot presently be settled. [38] If the latter proves true, the primary evidence for the "US–Israel rift trigger" argument discussed in §14 weakens — and we flag that explicitly.
The US leans toward "Iranian concession," the Iranian government toward "Iranian victory," the hardliners toward "Iranian capitulation." Net them out and the MOU looks, Lebanon aside (nearly the only square Iran actively secured), like a largely US-led asymmetric bargain. Each camp selling the same document differently to its domestic audience — this is the classic shape of constructive ambiguity, and the seed that bursts into interpretive conflict in the 60-day phase.
This is the heart of the document. Lay the MOU's 14 clauses in a row and they all look like "the agreement," but each clause is in fact poured into a completely different legal vessel. Some are enforced by a Security Council resolution; some are administrative acts of the US Treasury (OFAC); some have no force without an act of the US Congress; some are pure bilateral promises; and for some, an enforcement mechanism scarcely exists.
Why does this matter? Because almost everything Iran most wants sits in a vessel the UNSC cannot enforce, while the one thing the US most wants (nonproliferation) sits in the vessel the UNSC can. This asymmetry produces every tension of the 60-day talks and the Security Council drafting stage.
| MOU clause | Legal track | Enforcer / veto | Core issue |
|---|---|---|---|
| Ceasefire all fronts incl. Lebanon | ⑤ non-inst. | wholly dependent on US pressure on Israel | Israel is not a party → cannot be enforced. Top rupture point of the 60-day window |
| Non-interference / sovereignty | ④ bilateral | already in UN Charter Art. 2 — not a new duty | symbolic/political clause. UNSC only as preamble rhetoric |
| Lift naval blockade in 30 days | ④ bilateral | unilateral US performance (already begun) | nearly the only US-side step with a fixed date |
| US troop withdrawal from region | ④ bilateral | sovereign US military decision — UNSC cannot | "surrounding" distance undefined → dispute-prone |
| Hormuz reopening (management regime) | ④ bilat. ① partial | bilateral + (at final deal) UNSC text | cost structure of the "regime" undefined → hard to codify |
| Lift oil/petrochem sanctions + fund access | ② OFAC | US Treasury waiver — UNSC irrelevant | US domestic law. Cannot be lifted by a Council resolution |
| $300bn reconstruction plan | ⑤ non-inst. | US + GCC voluntarism — no enforcement | "asked, but they refused" (Nabavian) |
| 60-day talks → final deal (nuclear · sanctions) | ① UNSC | Council — Russia/China veto | strongest UNSC frame. Inverted veto is the variable |
| Reaffirm NPT non-weapon obligation | ① UNSC | can be codified as reaffirmation in a resolution | reaffirming an existing duty → legally clean |
| No force build-up / no new sanctions during talks | ③ Congress | no force without congressional involvement | collides with Trump's "re-attack if talks fail" |
| Release $24bn frozen funds (half upfront) | ② OFAC | US Treasury — license issuance | weakened to "based on progress." Amount/timing vague |
| Implementation-monitoring architecture | ① nuclear ④ rest | nuclear = IAEA / rest = mutual watch | only nuclear is institutionalized; rest has no clear arbiter (§12) |
| Final deal approved by UNSC resolution | ① UNSC | Council — Russia/China veto | repeats the 2231 structure. Inverted veto is key (§11) |
| Agenda limited (missiles/proxies excluded) | ① UNSC ⑤ pressure | an Iranian win, but Israeli lobbying is the variable | risk of missiles re-annexed to the resolution as in 2231 |
Read the table by color and it leaps out. Everything Iran is desperate for (OFAC sanctions relief, the $24bn, a no-re-attack guarantee, Lebanon) sits entirely in tracks ②④⑤ — vessels the Council cannot enforce. Meanwhile the one thing the US is desperate for (nonproliferation) sits in track ①, the UNSC — the vessel the Council can bind tightly. So Iran goes to the UNSC, but the UNSC can barely guarantee what Iran wants while binding tightly the nuclear concession Iran must give. This is why Gharibabadi's "active distrust" is no empty phrase, and why Nabavian's "the UNSC is no guarantee" is structurally correct.
The detail of the six prior sanctions resolutions restored in September 2025 is not mere history. It is the menu of "what to lift in a final agreement" and, conversely, "what can be plastered back on." The key is that the six are not one replacing another, but a cumulative structure stacked layer upon layer as Iran kept refusing to comply.
Note the 1737 Committee created by 1737. [36] A standing body that oversees implementation and adds new individuals and entities to the list, it was revived when the six were restored by snapback. In theory the engine that endlessly expands the list is running again. But Russia and China are blocking its ignition. The US Mission to the UN charged that Russia and China obstructing the chairing of the 1737 Committee and blocking its 90-day report is not a legal objection but a bid to protect their partner, Iran. [37] The engine is revived, but Russia and China hold it stalled.
Here lies a hidden bomb for the talks ahead. The six restored resolutions and the MOU's negotiating agenda overlap only partly.
| Domain | In the six? | On the MOU agenda? | Grounds to lift |
|---|---|---|---|
| Nuclear (enrichment/reprocessing) | Yes (1696·1737 etc.) | Yes (enrichment handling) | match → clear grounds to lift |
| Ballistic missiles | Yes (1929) | No (excluded) | divergence → pretext to retain |
| Conventional arms embargo | Yes (1929) | No (excluded) | divergence → pretext to retain |
| Asset freeze / travel ban | Yes (1737–1929) | Partly (sanctions relief) | partial match |
| Proxy (resistance) support | Indirect | No (excluded) | divergence |
Iran explicitly removed missiles and proxies from the MOU agenda. In the short run this looks like a win — "our missiles are not up for negotiation." But at the UNSC stage it becomes a boomerang. Iran's logic ("we negotiated only the nuclear file, so lift only the nuclear sanctions") hands the US and Israel the converse pretext — "then there's no reason to lift the missile and arms embargo." Excluded from the agenda, those very sanctions (the heart of 1929 among the six) lose the grounds for termination.
What Israel most fears about Iran is precisely its ballistic missiles and Hezbollah (proxies). So "terminate all six wholesale" is something the US and Israel can never easily accept. The most realistic path therefore becomes "lift only the nuclear portion, and attempt to re-package the missile/arms embargo in a new resolution — which then runs into a Russia/China veto." Iran pushes "lift both," the US and Israel push "keep missiles/arms," Russia and China push "lift everything (they want to sell arms to Iran)" — a three-way deadlock. This divergence is the fault line that erupts in §11 (UNSC process) and §13 (Israel risk).
Think of the 1737 Committee as an "engine," and its life cycle becomes the hidden clock of these talks. The engine is the standing body that keeps adding new individuals and entities to the sanctions list, so "can sanctions be plastered back on" is, in effect, the question of "is this engine running."
A striking contradiction is underway. The US State Department, while signing a peace MOU with Iran on one hand, is on the other hosting symposia to secure technical-assistance pledges from various countries to enforce the six restored resolutions more robustly. [35] This asymmetry — moving toward a nominal lifting while intensifying actual pressure — is the real-time evidence of "re-sanctioning risk": even when a UNSC "lifting" headline lands, real de-risking proceeds far less than the headline so long as US new-designation capacity stays alive.
The MOU draft states that "the final agreement will be approved by a UN Security Council resolution." [27] This is the core of track ① in the Table of §9, and the point at which the 2231 structure analyzed in §3–§6 re-activates in real time. Following the 2231 precedent, the process runs in five stages.
A common misconception must be dispelled precisely here. Even when the US and Iran sign a final agreement in Geneva, that signature alone lifts not a single one of the six restored resolutions. The final deal is a bilateral document, while the UN sanctions belong to the whole international community. They lift only if the Council actively passes a separate new resolution that "terminates the six."
Why not automatic? The answer lies in the structure of the resolutions themselves. The six originally have no sunset clauses — no "expires after X years" — and their structure reads: "if the Council judges that Iran has performed its obligations, the Council terminates the sanctions." The trigger for termination is not automatic but an affirmative Council decision. This is the exact mirror of the snapback logic in §3 — just as snapback required "an affirmative act (a yes-vote) to prevent," lifting requires "an affirmative act (passing a new resolution) to undo." Left alone, the six live forever.
So the actual sequence is: ① the US and Iran reach a final deal → ② someone (the US or E3) tables a new resolution that "endorses the deal and terminates the six prior sanctions" → ③ it is adopted with P5 non-veto + 9 affirmative votes → ④ only then do the six terminate. And this "terminate the six" resolution is negotiated as two parts fused into one. (A) The termination part is relatively easy to pass — Russia and China, holding that "the restoration was void," have no reason to oppose termination, and the US is willing to lift it as the price of the deal. (B) The new-snapback part is where it jams — even if the US and Israel want to "plant an automatic trigger to re-impose on breach," that is an ordinary resolution with no veto-proof exemption, so Russia and China can veto it.
Because this passage is decisive for reading the mood of the negotiation, it must be verified against reporting and official statements rather than asserted. The short answer: that the US does lift sanctions is confirmed, but the crucial caveat is that it is "conditional, traded for performance," not "unconditional."
The most authoritative official expression is the leaders' joint position on the day the deal was announced. They stated they "are prepared to lift relevant sanctions in response to clear, verifiable steps by Iran on its nuclear programme." [47] The phrase "in response to clear, verifiable steps" compresses the US position — it will lift, but only after Iran moves first.
The working-level expression is more concrete. A senior US official defined Trump's core principle as "relief for performance," stressing that "Iran wanted immediate and permanent sanctions relief, but that would only happen after tangible concessions were made." [48] The actual MOU draft, in exchange for reopening Hormuz and clearing mines, has the US lift the blockade and issue waivers letting Iran sell oil freely for 60 days, with relief increasing if Iran complies with the initial agreement and shows "good faith" in subsequent talks. [49]
① It treats sanctions relief as "leverage." The White House sees itself in a strong position and wants to tie economic relief not only to ending the nuclear program but also to curbing support for regional proxies — "if economic relief is tied to commitments to both end Iran's nuclear program and its support for terror proxies, then the president will have addressed the breadth of Iran's threats." [50] For the US, lifting sanctions is not a concession but a trading card to extract more.
② An exit from the war. Ending a war that dragged on for more than 100 days requires a "carrot" for Iran, the core of which is sanctions relief and access to frozen funds. Trump declaring the deal "complete" while leading with the blockade lift and Hormuz opening is part of this exit strategy. [51]
③ It controls risk through sequencing. By designing relief as "phased and conditional on performance" rather than "immediate and permanent," it preserves room to re-tighten if Iran breaks its promises. This is the revival of the 2013 JPOA principle of "performance first, money later" (§7).
One distinction matters, however. The "sanctions relief" in the reporting above mostly refers to US unilateral sanctions (OFAC waivers, track ② of §9), which the US executive can lift on its own, relatively quickly. By contrast, the subject of this section — UNSC termination of the six (track ①) — is a different order, requiring a Council resolution. So the favorable mood ("the US is willing to lift") is already being demonstrated on the OFAC track, but whether it carries smoothly into "termination (A)" of the UNSC six — and whether it collides with Russia and China over "the new snapback (B)" — is a separate question. The favorable mood (that the US intends to lift) and the structural obstacle (on which track, with what language, the lifting happens) exist simultaneously.
① The inversion of the veto configuration. In 2015 the veto-proof snapback was the West's weapon. But with 2231 gone since October 2025, the new resolution's restoration device follows ordinary procedure, so Russia and China hold the veto. In 2015 a "snapback for sunsets" bargain was possible (Obama–Lavrov, before the Ukraine war); today's Moscow has no reason to gift the West a veto-proof enforcement device again — certainly not for free. So the final resolution is likely to be "terminate sanctions, but with a weak or absent restoration device" — favorable to Iran, unfavorable to the US and Israel.
The mood does begin favorably. Italy, France, Germany, and the UK have signaled willingness to lift Iran sanctions after a deal. [28] That is, "(1) lifting is in everyone's interest," per the Table in §9, is falling into place. The problem is (2) the new enforcement architecture and (3) the re-sanctioning risk — and those burst at the final-deal drafting stage 60 days on.
② The politics of sequencing. Russia and China hold that "the 2025 snapback was itself void, so there are no sanctions to terminate in the first place." So the fight starts over whether to draft the text as "terminate" or merely "noting" — because voting to "lift" implicitly concedes the sanctions were valid. A drafting war precedes any vote.
③ Iran's "active distrust." Iran takes the UNSC endorsement but does not trust it as a guarantee. It will use the resolution not as an enforcement tool but as a device to raise the political cost of US re-defection. No one knows better than Iran what became of "a US president's promise" and Resolution 2231 in 2018.
The "six-vs-MOU agenda mismatch" seen in §10 (nuclear overlaps; missiles/arms diverge) produces a concrete branching here. The "terminate the six" resolution does not simply pass; it flows into one of the following three.
The conclusion running through all three: agreement on "terminating the six" is easy on the nuclear part but near-impossible on the missile/arms part. So the most realistic path is Scenario B, ending in "a patchwork partial termination plus permanent deadlock over missiles and arms." Iran's reason for going to the UNSC while holding "active distrust" deepens one more layer here — the UNSC may unlock the nuclear file, but nowhere is there a guarantee it will cleanly unlock missiles and arms too.
The analysis so far rests on the premise that "the final deal is approved by a UNSC resolution." That premise is not arbitrary — the MOU draft says so. [27] But when an analysis converges this neatly in one direction, intellectual honesty requires naming the path by which the premise itself could be bypassed. A non-Council route is structurally open.
The key is the asymmetry of the Table in §9. Most of the sanctions relief Iran actually feels comes from track ② (OFAC / US Treasury waivers) and EU autonomous de-listing, not from track ① (the UNSC). That makes the following sequence possible — while the 60-day talks stretch out under the empty verification baseline (§12), the US lifts oil and fund sanctions via OFAC licenses, the EU rolls back its own sanctions, and at the Council a mere "noting" presidential statement (or nothing at all) replaces a "terminate" resolution. The six restored resolutions then survive legally (Russia and China deem them void and don't enforce them anyway), while the real economic effect has already landed through the bilateral and EU channels.
Here Russia and China's calculus paradoxically aids the bypass. They hold that "the 2025 snapback was void, so the six never existed." They therefore have little active incentive to pass a 'termination' resolution at the Council — voting to terminate would implicitly concede that the restoration was valid, a self-contradiction (§11 ②). Leaving the resolution untabled and the six as "dead letters" is more consistent with their legal narrative. So an tacit collusion can form: the US banks the substance via bilateral/OFAC channels, Russia and China keep their "the sanctions were always void" line, and both sides avoid a Council vote. (There is speculation that Russia would demand Ukraine-linked concessions in exchange for veto restraint, but the Kremlin gives almost no public signals, so the exchange function remains a matter of conjecture — this document leaves it as an unresolved variable, not a settled analysis.)
The implication of this counter-case is heavy. If the final deal bypasses the Council, the entire "inverted veto → no new snapback → Scenario B" axis of this document does not operate. The six remain neither terminated nor re-packaged — "legal zombies" — and the substance of sanctions rests wholly on the administrative discretion of Washington and Brussels. In that case Iran's "active distrust" is even more justified — it receives no multilateral guarantee (UNSC) at all and must rely solely on bilateral administrative acts (OFAC waivers), which are exactly the kind of promise Trump overturned with a single executive order in 2018. In short, whether the deal actually goes through the Council is the single largest open branch in this negotiation, and the snapback-centered logic of this analysis must be read as conditional on "the Council actually being used."
The MOU writes a single line — "establish an architecture to monitor implementation." In reality that architecture has wildly different maturity by domain. Three layers operate at once, but asymmetrically.
Nuclear — the IAEA, but with an empty ledger. Disposing of HEU, freezing enrichment, and verifying sites fall to IAEA inspectors, cameras, and seals. But a decisive problem: in 2013 one could read the dials of operating facilities, whereas the 2025 bombing destroyed Fordow, Natanz, and Isfahan and Iran cut off access — so the IAEA must rebuild "what remains and how much" from scratch. [14] Trump himself conceded that the nuclear material was "buried under a mountain" after the strikes, beyond anyone's reach. [29] Building a new baseline atop ruins within 60 days is physically near-impossible. This is the weakest link in the architecture, and why a rollover of the 60-day window is the default rather than the exception.
Hormuz / blockade — mutual military watch. Verifying demining and monitoring transit takes the form of US Navy and UK/French minesweeping assets and Iran observing each other, with no neutral third-party verifier, so the arbiter of "implementation complete" is ambiguous.
Funds / sanctions — effectively unverifiable. Issuing OFAC licenses and activating payment channels are unilateral acts of the US Treasury, which Iran has no way to "monitor," and conversely the US judges Iran's performance. So the architecture is institutionalized only on the nuclear side (IAEA), with everything else a matter of reciprocal wariness.
So far we have analyzed the external actors (US, Russia, China, E3) and the Council. But the nearest hand that could actually kill this deal sits inside Tehran. Markets and foreign media play up the hardliners' street rallies and parliamentary outbursts, yet the question to ask is a different one — does that noise have the authority to actually block the deal, or only to embarrass? The two are entirely different, and that distinction is the substance of "Iran risk."
Iran's ratification path is clearly layered. The negotiating team (Araghchi, Ghalibaf) drafts the text, the Supreme National Security Council (SNSC) deliberates and approves, parliament reviews, and above it all sits the Supreme Leader's assent — the final arbiter of all state policy. The 2015 precedent shows the template — Ali Khamenei then permitted a parliamentary vote while keeping the final authority himself, saying it was "not for lawmakers to ratify or reject, but to review." [39] That is, parliament and the hardliners are instruments of delay, pressure, and public opinion, not in themselves a gate holding a veto.
That said, the hardliner noise should not be dismissed. Even with weak blocking power, it has two real effects. First, it raises the threshold for the Supreme Leader's assent — when protesters chant that "no understanding is valid without Mojtaba Khamenei's approval," they paradoxically push final responsibility onto the Supreme Leader, raising his political cost. [40] Second, erosion at the implementation stage — even if they cannot block ratification, branding the negotiating team "infiltrators" (Nabavian) and the extreme rhetoric of some MPs who say the talks would lead to the Supreme Leader's "assassination" [41] chip away at the will to implement.
Here lies the decisive variable. Supreme Leader Mojtaba Khamenei has barely appeared in public since his designation, and the hardliners are filling that silence. The hardliner-controlled Tehran municipal authority hung a giant banner declaring "the Strait of Hormuz will remain closed," as if signaling an edict from the new leader. [42] That is, with the legitimacy of a new power — hastily elevated after his father (Ali Khamenei) was assassinated — not yet consolidated, the hardliners exercise a proxy right to interpret "the Supreme Leader's true intent." This is the deepest uncertainty in Iran risk — the formal authority to block the deal rests with one man, that man is invisible, and his silence is being filled by the deal's opponents. While markets simplify this into "hardliners vs. negotiating team," the real question is "when, and in which direction, will the invisible Supreme Leader break his silence?"
The desk implication is clear. Over-reacting to hardliner headlines (rallies, signature refusals) is misreading noise as signal — since the SNSC has already confirmed the MOU, the center of gravity of ratification lies not with parliament but with the Supreme Leader's assent. Yet the fact that this assent is an unobservable black box is itself the deferred tail risk. The Supreme Leader's first public statement, the SNSC's formal ratification vote, and the Guardian Council's commencement of review — these three are far more important observation nodes than the parliamentary noise.
MOU clause 1 — "immediate and permanent end of military operations on all fronts, including Lebanon" — was sorted into track ⑤ (non-institutional) in the Table of §9. The reason is plain: Israel is not a party to this agreement. Netanyahu publicly declared that "Israel is not a party," and Israeli sources went so far as to signal that they "do not acknowledge that an agreement was reached." [30]
Israel is the only actor in this entire architecture that sits at no table yet wields the physical power to overturn every table. It is not an MOU party, was not a JCPOA participant, and is not a Council member. So there is no enforcement mechanism — the only way to enforce the Lebanon ceasefire is for the US to pressure Israel, a purely bilateral (US–Israel) process unrelated to the UNSC.
Here lies a structural paradox. Israel's traditional shield at the Council was the US veto. But the resolution coming up is one the US itself sponsors. For the first time, Israel faces "a US-pushed Iran resolution that no US veto can block." The remaining tool is to lobby inside Washington (Congress, hawks) at the drafting stage — to harden the new snapback trigger and tighten the conditions for relief.
A strange configuration results. Israel wants the enforcement device strong; Russia and China want it weak — and both push in a direction that makes the resolution harder to adopt. A spoiler coalition of opposite motives at the two ends. The force eating away at the multilateral guarantee Iran wants lies not only in Moscow but in Jerusalem.
Lebanon is no theoretical risk. On the very day the MOU signing was announced, Israel struck a Hezbollah position in southern Beirut, Iran threatened retaliation, and diplomacy briefly teetered toward derailment. [31] Since April the Israel–Hezbollah ceasefire has been a "ceasefire on paper," and on 6 June Israel struck a Lebanese army vehicle, killing three including a general. [32]
A US–Israel rift also lurks in the nuclear verification. Netanyahu's office announced that Trump had guaranteed the "removal" of enriched material. But the actual text reads "blend down all nuclear material on site, at minimum under IAEA supervision and under American considerations" — domestic dilution, not removal. What Trump sold Netanyahu differs from what the document says. The moment that gap surfaces during the 60-day window becomes the first trigger of a US–Israel rift. With the verification baseline empty, Israel's intelligence cards (HEU whereabouts, suspected covert sites) carry far more weight than in peacetime — exactly as the 2018 archive disclosure shaped the IAEA agenda for years.
The 19 June signing is not an end but the starting point at which the 2231 structure analyzed in §3–§6 — inverted veto, asymmetric withdrawal fear, re-sanctioning risk — begins operating in real time. The expected path, in chronological order.
The variable running through every phase is domestic politics on both sides. The Washington bottleneck (Congress, OFAC — legislating the permanent lift of statutory sanctions) and the Tehran bottleneck (hardliners — eroding ratification and implementation) both gnaw at the final deal at once. Both sides delegated the MOU stage to executive discretion, but the final-deal stage draws in both legislatures. That Trump declared a "re-attack if talks fail" stance, colliding head-on with the "no new sanctions / no force build-up during talks" clause, is part of this bottleneck.
A common market error is to read "signing = completion of the multilateral guarantee." In reality the hardest gates lie after signing — the multilateral gate (Moscow's veto), the physical gate (a verification baseline in ruins), and two domestic gates (Washington and Tehran). The keys to those gates lie not only in Washington and Tehran but in Moscow, and the hammer to break the locks lies in Jerusalem.
The machine that began in Geneva thirteen years ago worked once, broke in 2018, and in 2025 the P5 split apart over its ruins. In June 2026, in the same city, a new machine is being assembled. But this time it must be done without a veto-proof device, atop a verification ledger left blank by bombing, in the midst of the war in Ukraine, with hardliners in both houses shouting to tear up the document, and an uninvited guest standing outside the door with a hammer. The signing is not the end of all that, but the beginning.