The memorandum of understanding that the United States and Iran agreed in mid-June 2026 was a 14-point framework with a 60-day fuse. It paused the fighting, promised to reopen the Strait of Hormuz to commercial shipping, offered waivers on oil sanctions and, by Iran’s account, access to frozen funds, and committed both sides to negotiate a final deal. On Iran’s highly enriched uranium it said only that the parties would resolve its disposition, at a minimum by blending it down. On enrichment it promised further talks, and so left open the question that has defined every nuclear negotiation with Iran for twenty years: whether Tehran will accept “zero enrichment” as a condition for a deal.
The memorandum did not survive its own window. One round of high-level talks, at Bürgenstock in Switzerland in late June, set up working groups on the nuclear file and on monitoring. Then strikes resumed within a fortnight, Trump declared the memorandum “over” on July 8, and the 60-day deadline lapsed on August 17 with no deal.
As of October 2026 there is still no deal. In early October Vice President JD Vance called for a “meaningful” reduction in Iran’s enrichment capacity rather than zero, which looked like a softening of Washington’s line and a tacit admission of the problem with the zero-enrichment demand.
The public framing treats this as a familiar standoff: the US demands maximum concession, Iran demands maximum retention, and somewhere in the middle sits a deal both sides can live with. That framing misreads the demand, which cannot be bargained down the way a number can. No state with a mature, home-grown enrichment industry has ever negotiated away its enrichment; Iraq lost its programme only after military defeat and under Security Council compulsion. The technical reality of uranium enrichment explains why.
One machine for fuel and for bombs
A gas centrifuge does not know what it is being used for. Uranium hexafluoride gas enters the rotor, spins at supersonic speed, and the heavier U-238 isotopes migrate outward while the lighter U-235 concentrates near the axis. The enriched fraction is drawn off and fed into the next centrifuge, and the next, and the next. String enough of them together into a cascade and the U-235 concentration climbs: from the 0.7 percent found in natural uranium to the 3 to 5 percent used in power-reactor fuel (the JCPOA capped Iran at 3.67), to the 20 percent used in research reactors and medical isotope production, to the 90 percent that constitutes weapons-grade material.
There is no switch labelled “civilian” and another labelled “weapons.” The enrichment level is a continuous variable. The same cascade that produces reactor fuel produces bomb material if you add stages or recycle the output as feedstock. A cascade configured for 3.67 percent can be turned toward weapons production by rearranging how the gas flows through equipment that already exists. How long that takes depends on the configuration and the feedstock, but no new kind of machine is needed.
By June 2025, Iran had accumulated 440.9 kilograms of uranium enriched to 60 percent, a short technical step from weapons grade. No other non-nuclear-weapon state has produced more than laboratory traces at that level. After the US raid of June 22, 2025, General Dan Caine, chairman of the Joint Chiefs of Staff, said all three major sites, Natanz, Fordow and Isfahan, had sustained “extremely severe damage and destruction.” Independent analysts judged that the Israeli and US attacks had effectively destroyed Iran’s centrifuge enrichment programme: most of the centrifuges at Natanz eliminated or badly damaged, Fordow’s halls hit by twelve GBU-57 Massive Ordnance Penetrators, the uranium conversion facility at Isfahan wrecked.
Bombs can wreck buildings and machines, but neither they nor a zero-enrichment demand can remove what Iran’s engineers know. The June 2025 strikes killed nearly twenty senior nuclear scientists, but most of Iran’s nuclear engineering corps survived, and so did its centrifuge design files, its materials science expertise and its manufacturing know-how. Over two decades, Iran designed and built successive generations of centrifuges itself, from the early IR-1, based on designs originally acquired through the A.Q. Khan network, to the IR-6, which does roughly five to ten times the enrichment work of an IR-1. That capability lives in the heads of engineers and in the institutional memory of the programme that trained them.
The centrifuge generations: from rented parts to indigenous capability
Iran's centrifuge programme began with stolen blueprints. The IR-1 was a copy of the Pakistani P-1, itself derived from early Dutch designs that A.Q. Khan took from the URENCO enrichment consortium in the Netherlands in the 1970s. It was crude, unreliable, and broke down constantly. But Iran did not stop at copying. Over two decades, Iranian engineers designed the IR-2m, the IR-4, the IR-5 and the IR-6, a machine so far from the original P-1 that calling it a descendant is generous. Iran manufactured these machines domestically, in workshops IAEA inspectors have not been able to enter since February 2021. This progression from purchased components to indigenous design and manufacturing separates Iran from Libya, which bought P-1 parts from the Khan network but never mastered centrifuge engineering. It is also why dismantling hardware does not amount to disarmament.Some countries have given up nuclear weapons or dismantled nuclear programmes. South Africa built six warheads and voluntarily dismantled them in 1990–91, just before joining the NPT, as apartheid was ending. Libya surrendered its centrifuge components in 2003. But neither case fits the pattern the zero-enrichment demand assumes. Libya never mastered centrifuge engineering; it bought parts off the Khan network’s shelf. South Africa’s programme did not rest on centrifuges at all, and it gave up its weapons voluntarily during a domestic political transformation, not under arms control pressure from an adversary.
The category that matters is states with an indigenous, mature enrichment and centrifuge-manufacturing capability, whose own scientists designed, built and operated their own cascades. No state in that category has ever accepted a demand to surrender the capability, whether under the NPT, under the JCPOA or at any negotiating table since 1945. Iran is in that category, and zero enrichment asks it to leave.
What arms control has done
The agreements that have constrained working enrichment programmes were built around this physics rather than against it. The model that works is threshold-plus-monitoring: you don’t try to eliminate the capability, because the know-how cannot be taken back. Instead, you cap what a state can do with it, in enrichment levels, stockpile sizes and centrifuge numbers, and you build a monitoring architecture invasive enough to catch violations before they become weapons.
The JCPOA was the most ambitious version of this model ever constructed. Its numerical limits were thresholds: a 3.67 percent enrichment ceiling, a 300-kilogram limit on the stockpile of low-enriched uranium hexafluoride (about 202 kilograms of uranium itself), and just over 5,000 IR-1 centrifuges enriching uranium, all at Natanz, with about a thousand more kept at Fordow for non-uranium work and the surplus in storage under IAEA seal. It also redesigned facilities: Fordow was converted from an enrichment facility to a research centre with no fissile material permitted, and the Arak heavy-water reactor was redesigned to produce minimal plutonium. Every one of these provisions assumed Iran would keep its enrichment capability. They were written to keep Iran about a year away from enough material for a bomb, and to make it very likely that any move in that direction would be spotted in time to respond.
The monitoring went further than anything before it. Surveillance cameras ran continuously at every declared enrichment facility. Online enrichment monitors reported in near-real time what was being produced and at what level. Monitoring reached back along the centrifuge production chain to the components used to build the machines, from maraging steel tubes to carbon fibre rotors, tracked from manufacturing floor to installation. What the JCPOA gave inspectors, in the jargon of nuclear safeguards, was “continuity of knowledge”: an unbroken record of Iran’s declared material and equipment, with no gaps in which something could be moved or added unseen.
That continuity eroded in stages. In February 2021, under a law passed in response to US “maximum pressure” sanctions, Iran stopped applying the Additional Protocol and the JCPOA’s extra verification measures. Complementary access to undeclared sites ended, and so did inspectors’ visits to the workshops where centrifuges were built.
Then the June 2025 military strikes shattered what remained. The IAEA pulled its inspectors out for safety, and in early July Iran’s president signed a law suspending cooperation with the Agency. A September 2025 agreement signed in Cairo laid out procedures for inspections at all declared facilities, but it never reached the bombed sites. Iran declared it terminated in November, after UN sanctions had snapped back at the end of September. Inspectors kept visiting the undamaged reactors and laboratories into February 2026, but they never got back into Natanz, Fordow or the Isfahan tunnels where the highly enriched uranium was stored. When the second war began on February 28, 2026, the IAEA withdrew its inspectors again.
The IAEA’s Additional Protocol, the enhanced inspection framework Iran accepted under the JCPOA, grants inspectors complementary access to any location where undeclared nuclear activities are indicated. They can take environmental samples, visit uranium mines and milling operations, and inspect centrifuge manufacturing plants. But even the Additional Protocol has limits. It is the tool the IAEA needs before it can give “credible assurance about the absence of undeclared nuclear material and activities,” and that assurance is a judgement built up over years of access, not proof. In a country the size of Iran, with an advanced manufacturing base, a small hidden workshop can still slip through.
Throughout the 1980s, IAEA inspectors verified Iraq’s declared nuclear facilities and found no diversions. Every safeguards report came back clean. Meanwhile, Iraq ran a parallel clandestine weapons programme, a Manhattan Project-scale operation employing thousands of scientists across multiple undeclared sites. It wasn’t exposed until after the Gulf War, when inspectors with post-conflict access walked into facilities they’d never been told existed and found industrial-scale enrichment equipment. It turned out that safeguards had only ever looked at what Iraq chose to declare, and some of the secret sites stood just across a berm from buildings inspectors had visited. The Iraq experience is the reason the Additional Protocol exists.
North Korea exposed the gap from another side. In 1992 the IAEA’s analysis of the plutonium and waste Pyongyang presented showed several reprocessing campaigns, not the single one it had declared, so more plutonium had been separated than it admitted. To find out how much, the agency asked for a special inspection of two undeclared nuclear waste sites at Yongbyon, and Pyongyang refused. North Korea responded by announcing its withdrawal from the NPT in March 1993, suspending it one day before it took effect, then expelling inspectors in December 2002 and walking out of the treaty for good in January 2003. In neither case did the IAEA fail to monitor what it could see; in both, the answers lay outside the declared sites.
The knowledge that won't stay in the box
In the early 2000s, US and British intelligence, and then IAEA investigators, pieced together the A.Q. Khan network. Khan, the father of Pakistan's bomb, had sold centrifuge designs and components to Libya, Iran, and North Korea. Libya handed everything over: centrifuges, components and weapon-design documents were airlifted to the United States in 2004. Iran used them as a starting point and built something better. North Korea used them to add a uranium-enrichment programme to its plutonium bomb. The Khan network demonstrated a principle that zero enrichment cannot overcome: once nuclear knowledge has been transferred, no amount of hardware dismantlement recalls it. Khan provided blueprints for P-1 centrifuges and, in Libya's case, even a Chinese-origin nuclear weapon design. The hardware was interdicted and hauled away, but the knowledge stayed loose. The zero-enrichment demand treats capability as something stored in facilities. The Khan network showed that it is stored in people, in documents and in institutional memory, and that it proliferates even when the physical programme is shut down.The JCPOA left Iran enriching, and zero would not have spared the inspectors their hardest job. A halt at declared facilities can be verified: in 2004 and 2005 the IAEA sealed and watched Iran’s centrifuge halls and components while Tehran observed a voluntary suspension, and Iran ended that suspension when it chose to. You can verify that a country is enriching below a certain threshold at declared facilities, or not enriching there at all. What no inspection can prove is that a country with thousands of trained nuclear engineers and decades of centrifuge experience has stopped enriching everywhere, forever; the most the IAEA can offer is an assurance that has to be renewed year after year, for as long as its access lasts. The JCPOA’s designers knew this. They built the most intrusive monitoring regime ever imposed on a functioning nuclear programme, and even it could only verify what it could see.
If a demand for zero enrichment were somehow accepted on paper, it would leave behind the same verification problem, because the risk that matters is a hidden workshop, and that risk is much the same whether the declared halls run at 3.67 percent or stand idle. Threshold-plus-monitoring works because it verifies something measurable. Zero enrichment asks inspectors to certify a permanent absence, when all they can offer is continuing assurance for as long as they are allowed to look.
What the war could not change
Two rounds of bombing did not solve the nuclear problem, though what they achieved was real. In June 2025, Israeli and US strikes wrecked most of the centrifuges at Natanz, and B-2 bombers hit Fordow, which is buried inside a mountain near Qom, with the heaviest conventional bombs in the US arsenal.
The second campaign, from February 28 to the April 7–8 ceasefire, went after what was left: the tunnel entrances to Natanz’s underground halls, the Arak heavy-water plant, the Ardakan yellowcake plant and a string of weapons-research sites. Satellite analysts afterward found no sign that Iran had restarted enrichment.
Satellite imagery showed that before the second war Iran had thrown a makeshift roof over the bombed-out pilot enrichment plant at Natanz, most likely to hide salvage work, and had reopened the northern entrance to the Isfahan tunnels. After Operation Midnight Hammer, the Pentagon estimated the setback at approximately two years. A capability that lives in human capital regenerates unless the humans themselves are removed, and the June 2025 killings thinned that capital without erasing it.
The second war also finished off what was left of the monitoring of Iran’s enrichment programme. On February 27, 2026, in its last assessment before the second war, the IAEA could do no better than repeat the Agency’s June 2025 estimate of 440.9 kilograms of 60 percent enriched uranium. It could no longer tell where that stockpile was, how big it was now, or how many centrifuges Iran had, and it warned of a “loss of continuity of knowledge over all previously declared nuclear material at affected facilities.” The next day the bombs fell again, and the inspectors left. By June 2026 they were back only for routine checks at the Bushehr power plant; the enrichment sites and the enriched uranium stayed out of reach.
Any new agreement now starts from a position that the JCPOA never had to contend with: inspectors who do not know how much enriched uranium exists or where it is. The baseline that took years to build is gone. Rebuilding it requires Iranian cooperation that doesn’t currently exist, and it requires starting from scratch in a country whose facilities have been bombed and whose monitoring infrastructure has been dark for months.
The stockpile itself is a mystery. Iran appears to have moved much of its highly enriched uranium into underground tunnels at Isfahan before the June 2025 Israeli strikes, Operation Rising Lion. Grossi, the IAEA’s chief, has said the agency believes a bit more than 200 kilograms of 60 percent enriched uranium is stored there. The pre-strike total was enough, if further enriched to weapons grade, for eight to ten bombs. The agency believes some of the rest is still at Natanz, and that some held elsewhere may have been destroyed in the June 2025 strikes. Nobody can confirm it. The monitoring blackout means the IAEA’s own figures are estimates, not verified accounts.
Two operations, one problem
The phrase "the US airstrikes" is shorthand for a two-phase, two-country military operation whose effects are still contested. In June 2025, Israel launched Operation Rising Lion, a campaign of strikes on Iranian nuclear and military infrastructure. US forces joined nine days later, on June 22, under Operation Midnight Hammer, striking Natanz, Fordow, and Isfahan directly. Iran appears to have anticipated the assault and transferred its most sensitive nuclear material to Isfahan's underground tunnel complex before the operation began. Then, from February 28 to the April 2026 ceasefire, a second US-Israeli campaign hit Natanz again and spread out across Iran's wider nuclear and missile-research network. Much of the material the US has demanded Iran surrender appears to have been moved to Isfahan because of the military operations that preceded the demand, and the June 2025 strikes did not fully destroy it.Iran’s politics have been remade too. Ali Khamenei was killed in the February 28 strikes. His son Mojtaba was named supreme leader by the Assembly of Experts on March 8, with the Revolutionary Guards quick to pledge their backing. In May, Reuters reported, citing two senior Iranian sources, that he had ordered that Iran’s enriched uranium must not leave the country, a report one Iranian official dismissed as propaganda. For a new supreme leader who came to power with the Guards behind him, after his father was killed by the countries now demanding he surrender the nuclear programme, there is little political space for capitulation.
The MOU’s nuclear section reflected all of this by deferring all of it. The US slogan was “no dust, no dollars”: Iran hands over its highly enriched uranium before any economic relief flows. That was an opening position for the 60-day window, never a settled term, and the window closed before anyone tested it. Where the stockpile would go was left open: out of Iran, which Mojtaba Khamenei has reportedly ruled out, or blended down inside the country, the floor the memorandum set. So was everything about an enrichment suspension: how it would work, how long it would last, who would verify it, and what would happen when a future US president decided the deal wasn’t working.
The 60-day clock ran out on August 17 with the enrichment and stockpile questions unsettled. The verification question, the one that determines whether any deal can hold, got a working group at Bürgenstock and nothing more.
The argument over years
What the two sides argued about, in the spring of 2026, was a number. The US proposed a twenty-year enrichment moratorium. Iran countered with three to five years. Trump at first rejected his own team’s proposal. “I don’t like the 20 years,” he said, before indicating he could accept it with “real” guarantees.
Twenty years against five at most sounds like a gap the two sides could split. Each number, though, rests on a different theory.
The US number makes sense if a dormant programme degrades over two decades, so that twenty years of zero enrichment erodes the institutional knowledge and manufacturing base enough to extend breakout timelines permanently. There is a logic to this: engineers retire, tacit knowledge fades, manufacturing equipment rusts. But the theory assumes that enrichment capability is stored primarily in hardware rather than in people and institutions, and the previous sections have shown that to be false. Iran reconstituted centrifuge manufacturing after setbacks before. The degradation theory also assumes that the commitment itself will hold for twenty years, which brings up the second problem.
Iran’s number fits a darker reading, that any commitment longer than the shortest political cycle is a fiction, and Iran has specific, recent evidence for that view: the Trump administration walked out of the JCPOA in May 2018 by executive decision, without a new Iranian violation, barely two and a half years after the deal was adopted. There was no congressional vote and no trigger clause. One president committed the United States; the next uncommitted it. Under US law, a president who can make a nonbinding political commitment like the JCPOA on his own authority can also drop it on his own authority, and that holds whether the commitment was meant to last five years or twenty. Treaty ratification, which requires two-thirds of the senators present, is politically unreachable for any Iran deal in the current or any foreseeable Senate. Neither duration comes with a guarantee the US system can provide.
That is how the last nuclear deal died. The duration fight also misses a larger point: a twenty-year moratorium with weak verification is worth less than a five-year moratorium with strong verification.
Duration tells you how long a commitment is supposed to last; that is a political promise. Verification tells you whether cheating will be caught; much of that is an engineering problem, and the engineering has solutions. Those solutions would make the moratorium’s length secondary, because cheating would stand a high chance of being caught in time to answer, whatever either side promised on paper.
Neither side answered the verification question. It was raised: Grossi sat in on two rounds of talks in February 2026 to advise on verification, and the June summit gave monitoring its own working group. But no workable answer made it into anything the two sides agreed, and it is easy to see why: answering it honestly would require each to name what its own political system cannot deliver. The US would have to explain how any agreement survives the next election, and Iran would have to explain how the IRGC honours what the Foreign Ministry signs. So both argued about numbers instead.
What Tehran cannot deliver
Iran’s nuclear programme has also been a sovereignty argument, built deliberately over three decades. The domestic case for enrichment rested on a specific claim: that nuclear energy dependent on Western-supplied fuel would leave Iran hostage to the same sanctions that had already crippled its economy. The elder Khamenei made this explicit in a speech on June 4, 2025, nine months before his death: “If we have a hundred nuclear power plants but no enrichment capability, they are worthless … If we don’t enrich ourselves, we’ll have to extend a begging hand to America — and they will impose conditions.”
That line has carried the political case for the programme for thirty years, and because it rests on a real sanctions history, any successor in Tehran inherits it along with the office.
The sovereignty narrative was designed to make a specific capability politically untouchable. Zero enrichment would make Iran give up both the capability and the argument that justified building it in the first place. Iran has accepted limits before: it suspended enrichment from late 2003 until early 2006 and kept to the JCPOA’s caps for years after 2016. Giving up enrichment for good is a different order of concession, and no political actor in Tehran could make it and expect to remain in power, whatever the temperament of Iranian leaders, because the programme was constructed specifically to make that concession politically impossible.
Mojtaba Khamenei’s situation makes this worse, and the reason lies in the structure around him. He came to power with the backing of the Revolutionary Guards, and he is widely seen as owing his position to them. His father, the previous supreme leader, was killed by the United States and Israel, the countries now demanding that Iran give up enrichment. A leader in that position has less room to concede on the programme the Guards consider their deterrent than any of his predecessors had, and the same would hold for anyone in that office under these conditions.
The guards and the programme
"Iran's position" on enrichment is not a single position. The Revolutionary Guards guard the programme, and they regard it as a strategic deterrent, not just an energy project. In March 2022, after a run of sabotage attacks, the IRGC set up a dedicated "Command for the Protection and Security of Nuclear Centres," taking over the security of the sites and formalising its institutional stake. The gap between the civilian government's negotiating stance and the Guards' institutional interest is structural. Any agreement the civilian government signs must also be honoured by an organisation whose interests, budget, and political identity are bound to the programme's continuation. Iranian "compliance" therefore turns on more than whether a president or foreign minister means what he says. It depends on whether the institutions that guard the programme will follow through on commitments made by the institutions that negotiate.Tehran also reads the Libya lesson the opposite way from Washington. US and Israeli officials have repeatedly cited Libya’s 2003 disarmament as the model for what Iran should do. Gaddafi surrendered his nuclear and chemical weapons programmes. The West welcomed him back, trade resumed and diplomatic relations normalised. Eight years later, NATO warplanes enforced a no-fly zone over Libya, rebels overran Tripoli, and Gaddafi was dragged from a drainage pipe and killed on camera.
Iranian officials have been citing that sequence ever since 2011, as a case study in what happens when you give up the only thing that makes attacking you costly. Tehran’s calculus today is not Tripoli’s in 2003, but other states weighing the bomb, North Korea most openly, have pointed to Libya’s fate as evidence against disarming.
The 2026 war reinforced that lesson: the country that disarmed in 2003 was overrun within a decade, while the country that didn’t was attacked twice, and its nuclear know-how survived both rounds; the second never touched the Isfahan tunnels where much of its uranium is believed to lie.
Signing an agreement and complying with it are different things. Temporary limits that leave the programme standing are one thing; a permanent end to enrichment is another, and on that the Iranian political system is built to refuse. The programme is wired into the regime’s founding claims, reinforced by thirty years of political investment, and now cemented by a war that killed the previous supreme leader. A demand for zero enrichment cannot achieve the US goal even if an Iranian negotiator signs a piece of paper granting it.
The narrower problem
If zero enrichment is unachievable and the moratorium dispute is a proxy fight, an agreement that holds would have to follow a design that has been on the table for years.
Iran’s enrichment knowledge cannot be taken away; Iran has had it for decades. What a deal can do is make sure that if Iran enriches beyond agreed levels or at undeclared facilities, it gets caught fast enough for the response to be credible. Its negotiators would have to accept that some risk of cheating remains and make the risk of detection high enough to deter, with breakout timelines long enough that detection triggers a response before weapons-grade material is produced in quantity.
The verification architecture this requires has three components: access, continuity, and speed. Access means the IAEA Additional Protocol in full force, going beyond basic safeguards over declared facilities to complementary access wherever undeclared activities are suspected, backed by satellite intelligence that the US and its allies would share with IAEA inspectors. Continuity means restoring the baseline destroyed by the 2025–2026 monitoring collapse: surveillance cameras, centrifuge production chain monitoring and material accountancy rebuilt from scratch. Speed means challenge inspection rights with response times measured in hours, instead of the JCPOA’s dispute process for contested access requests, which could run for up to twenty-four days, a window arms control critics argued was wide enough to conceal small-scale violations.
A deal also needs consequences for cheating. Resolution 2231’s snapback mechanism was designed to be veto-proof. It inverted normal Security Council procedure so that sanctions relief required an affirmative vote to continue, meaning any single P5 member could block continuation and trigger automatic reimposition. The E3 invoked it in late August 2025. A Security Council draft to keep sanctions relief alive failed on September 19, Russian and Chinese efforts to delay the process went nowhere, and the old UN sanctions returned on September 28, along with the old resolutions’ demand that Iran suspend all enrichment. The legal mechanism worked, but enforcement lacked political credibility: Iran rejected the snapback as “null and void,” and the United States, the party that had walked out of the JCPOA in 2018, was in no position to invoke a deal it had abandoned. Any new consequence mechanism needs to survive what the last one didn’t: a change of US administration.
The US exit from the JCPOA did not shut the inspectors out at once, but it set off the chain that did, and the monitoring it took years to build now has to be rebuilt from scratch, assuming Iran cooperates, which it has little reason to do after being abandoned by the party that demanded the monitoring in the first place. The IAEA’s current blindness on Iran began with the 2018 withdrawal. Each political rupture since then stripped away another layer of verification, until the Agency could no longer see Iran’s enrichment programme at all. Building a new agreement requires building a new monitoring baseline. It also requires a way for that baseline to survive a future US president who decides the deal isn’t working, and nobody has found one yet.
Any durable deal also requires Israel not to bomb its way out of it the moment it perceives a threshold being crossed. Israel has shown that it will strike Iranian nuclear infrastructure on its own, as it did in June 2025, and alongside the United States, as in February 2026. An agreement that does not include Israeli acquiescence lasts until the next Israeli strategic assessment, and no longer. One Israeli analyst, Danny Citrinowicz, argues that an agreement, not another war, is what best protects Israel, and that Israel must back one if it wants its military gains to last.
The technical elements of a workable verification architecture are well understood: centrifuge restrictions on advanced models, stockpile limits below the weapons-material threshold, and the monitoring architecture described above, all of them solvable engineering problems.
The political elements are harder: a US commitment that survives a change of president, an Israel that accepts an agreement permitting some enrichment and holds its fire while the agreement holds, and an IRGC that honours commitments made by the civilians who negotiate. These are what the 60-day clock ran out on. None of them appeared in the MOU.
The 60 days exposed these obstacles rather than resolving them. The clock has already run out once, and when it did, the question of how to verify any deal had got no further than a working group.
Gen AI Disclaimer
Some contents of this page were generated and/or edited with the help of a Generative AI.
Media
Key Sources and References
IAEA Board of Governors, “Implementation of the NPT Safeguards Agreement and Relevant Provisions of United Nations Security Council Resolutions in the Islamic Republic of Iran,” GOV/2026/8, February 27, 2026. https://www.iaea.org/sites/default/files/gov2026-8.pdf
IAEA Board of Governors, “Implementation of the NPT Safeguards Agreement and Relevant Provisions of United Nations Security Council Resolutions in the Islamic Republic of Iran,” GOV/2026/33, June 4, 2026. https://www.iaea.org/sites/default/files/gov2026-33.pdf
IAEA Board of Governors, “Implementation of the NPT Safeguards Agreement and Relevant Provisions of United Nations Security Council Resolutions in the Islamic Republic of Iran,” GOV/2026/50, September 1, 2026. https://www.iaea.org/sites/default/files/gov2026-50.pdf
Swiss Federal Department of Foreign Affairs, “Memorandum of Understanding between the USA and Iran,” June 22, 2026 (Bürgenstock talks). https://www.fdfa.admin.ch/en/memorandum-of-understanding-between-the-usa-and-iran
Ministry of Foreign Affairs of Qatar, “Joint Statement by the State of Qatar and the Islamic Republic of Pakistan Regarding the Conclusion of Lake Lucerne Summit,” June 22, 2026. https://mofa.gov.qa/en/latest-articles/statements/joint-statement-by-the-state-of-qatar-and-the-islamic-republic-of-pakistan-regarding-the-conclusion-of-lake-lucerne-summit–first-high-level-committee-meeting-with-participation-of-the-united-states-of-america-and-the-islamic-republic-of-iran
ABC News, “Trump says MOU is ‘over’, calls Iranian leaders ‘scum’ following latest strikes,” July 8, 2026. https://abcnews.com/amp/International/trump-mou-calls-iranian-leaders-scum-latest-strikes/story?id=134576539
Institute for Science and International Security, “Significance of the Targeted Nuclear Scientists in the 12-Day War,” July 15, 2025. https://isis-online.org/isis-reports/significance-of-the-targeted-nuclear-scientists-in-the-12-day-war
U.S. National Archives, “The Constitution of the United States: A Transcription” (Article II, Section 2). https://www.archives.gov/founding-docs/constitution-transcript
Congressional Research Service, Stephen P. Mulligan, “International Law and Agreements: Their Effect upon U.S. Law,” RL32528, September 19, 2018 (section on withdrawal from executive agreements and political commitments). https://www.congress.gov/crs_external_products/RL/PDF/RL32528/RL32528.20.pdf
Institute for Science and International Security, “Post-Attack Assessment of the First 12 Days of Israeli and U.S. Strikes on Iranian Nuclear Facilities,” June 24, 2025. https://isis-online.org/isis-reports/post-attack-assessment-of-the-first-12-days-of-israeli-strikes-on-iranian-nuclear-facilities
Institute for Science and International Security, “Comprehensive Analysis of Nuclear Facilities Targeted During the Second Phase of the Iran War,” May 7, 2026. https://isis-online.org/isis-reports/comprehensive-analysis-of-nuclear-facilities-targeted-during-the-second-phase-of-the-iran-war
Institute for Science and International Security, “A Comprehensive Survey of Iran’s Advanced Centrifuges.” https://isis-online.org/isis-reports/a-comprehensive-survey-of-irans-advanced-centrifuges
Fox News, “Iran’s Fordow nuclear site struck second time as IDF targets access routes,” June 2025 (includes Gen. Dan Caine’s June 22, 2025 damage assessment). https://www.foxnews.com/world/irans-fordow-nuclear-site-struck-second-time-idf-targets-access-routes.amp
Al Jazeera, “US says its strikes degraded Iran’s nuclear programme by one to two years,” July 3, 2025. https://www.aljazeera.com/news/2025/7/3/us-says-its-strikes-degraded-irans-nuclear-programme-by-one-to-two-years
Bulletin of the Atomic Scientists, “Analysis: Iran likely transferred highly enriched uranium to Isfahan before the June strikes,” March 29, 2026. https://thebulletin.org/2026/03/analysis-iran-likely-transferred-highly-enriched-uranium-to-isfahan-before-the-june-strikes/
Foreign Policy, “IAEA’s Grossi: Much of Iran’s Enriched Uranium Likely Still at Isfahan,” April 29, 2026. https://foreignpolicy.com/2026/04/29/iran-nuclear-program-iaea-grossi-isfahan-enriched-uranium-us-war-hegseth-caine/
Al-Monitor (Reuters), “Much of Iran’s near-bomb-grade uranium likely to be in Isfahan, IAEA’s Grossi says,” March 2026. https://al-monitor.com/originals/2026/03/much-irans-near-bomb-grade-uranium-likely-be-isfahan-iaeas-grossi-says
IAEA Board of Governors, “Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran,” GOV/2005/67, September 2, 2005 (verification of Iran’s enrichment suspension). https://www.iaea.org/sites/default/files/documents/gov2005-67.pdf
IAEA Board of Governors, “Implementation of the NPT Safeguards Agreement in the Republic of Korea,” GOV/2004/84, November 11, 2004. https://www.iaea.org/sites/default/files/gov2004-84.pdf
IAEA, “The DPRK’s Violation of its NPT Safeguards Agreement with the IAEA,” excerpt from David Fischer, History of the International Atomic Energy Agency (IAEA, 1997). https://www.iaea.org/sites/default/files/dprk.pdf
Israel National News, report on IAEA inspectors’ withdrawal from Iran after the renewed attacks (relaying Iran International), 2026. https://www.israelnationalnews.com/news/433134
Arms Control Association, “Restoring the JCPOA’s Nuclear Limits,” factsheet. https://www.armscontrol.org/factsheets/restoring-jcpoas-nuclear-limits
Arms Control Association, “The Joint Comprehensive Plan of Action (JCPOA) at a Glance,” factsheet. https://www.armscontrol.org/factsheets/JCPOA-at-a-glance
Arms Control Association, “IAEA Verification of the Iran Nuclear Deal: 10 Hot Issues,” July 2015. https://www.armscontrol.org/blog/2015-07-23/iaea-verification-iran-nuclear-deal-10-issues
Arms Control Association, “Looking Back: The Additional Protocol,” November 2007. https://www.armscontrol.org/act/2007-11/looking-back-additional-protocol
Arms Control Association, “Chronology of U.S.-North Korean Nuclear and Missile Diplomacy,” factsheet. https://www.armscontrol.org/factsheets/dprkchron
Arms Control Association, “Chronology of Libya’s Disarmament and Relations with the United States,” factsheet. https://www.armscontrol.org/factsheets/LibyaChronology
Federation of American Scientists, “South Africa Nuclear Weapons.” https://nuke.fas.org/guide/rsa/nuke/index.html
Security Council Report, “Iran: Vote on a Draft Resolution regarding the ‘Snapback’ of UN Sanctions,” September 2025. https://www.securitycouncilreport.org/whatsinblue/2025/09/iran-vote-on-a-draft-resolution-regarding-the-snapback-of-un-sanctions.php
CBS News, “Here’s how Trump’s memo of understanding with Iran compares to the Obama nuclear deal,” June 17, 2026. https://www.cbsnews.com/news/how-trump-iran-nuclear-deal-memorandum-of-understanding-compares-to-obama-nuclear-deal-jcpoa/
Al Jazeera, “US-Iran Memorandum of Understanding expires: How and why it fell apart,” August 17, 2026. https://www.aljazeera.com/news/2026/8/17/us-iran-memorandum-of-understanding-expires-how-and-why-it-fell-apart
Al Jazeera, “US sets new demands for Iran deal: What are they?”, October 7, 2026. https://www.aljazeera.com/news/2026/10/7/us-sets-new-demands-for-iran-deal-what-are-they
Annahar, “No dust, no dollars: Slogans and strategy in the Iran standoff,” May 27, 2026. https://en.annahar.com/en/313899/trump-and-iran-on-the-slogan-no-dust-no-dollars
Newser, “Trump says Iran talks could restart within 2 days,” April 14, 2026 (Trump’s remarks to the New York Post; Iran’s three-to-five-year counter-offer). https://www.newser.com/story/387330/trump-says-iran-talks-could-restart-within-2-days.html
Times of Israel, “Trump says he’s OK with Iran suspending nuke enrichment for 20 years if there’s ‘real’ guarantee,” May 15, 2026. https://www.timesofisrael.com/trump-says-hes-ok-with-iran-suspending-nuke-program-for-20-years-if-theres-a-real-commitment/
Ukrinform, “Trump sets timeframe for moratorium on Iran’s nuclear program,” May 2026. https://ukrinform.net/rubric-polytics/4123923-trump-sets-timeframe-for-moratorium-on-irans-nuclear-program.html
Times of Israel, “Trump’s emerging Iran deal forces Israel to seek guarantees, not victory,” May 27, 2026. https://www.timesofisrael.com/trumps-emerging-iran-deal-forces-israel-to-seek-guarantees-not-victory/
NIAC, “Khamenei: ‘Who are you to decide whether Iran should enrich uranium or not?'”, June 5, 2025. https://niacouncil.org/khamenei-who-are-you-to-decide-whether-iran-should-enrich-uranium-or-not/
Al Jazeera, “Iran names Mojtaba Khamenei as new supreme leader after father’s killing,” March 8, 2026. https://www.aljazeera.com/news/2026/3/8/iran-names-khameneis-son-as-new-supreme-leader-after-fathers-killing-2
Khaleej Times (Reuters), “Khamenei says highly enriched uranium must stay in Iran, say two senior sources,” May 2026. https://www.khaleejtimes.com/world/mena/khamenei-highly-enriched-uranium-stay-iran
Atalayar, Cyrus Yaqubi, “Nuclear Corps Command is a step towards further secrecy in the Iran Nuclear program,” March 24, 2022. https://www.atalayar.com/en/opinion/cyrus-yaqubi/nuclear-corps-command-step-towards-further-secrecy-iran-nuclear-program/20220324170113136325.html
Atlantic Council, “From Tripoli to Tehran: Lessons from Libya in US-Iran nuclear talks,” April 30, 2025. https://www.atlanticcouncil.org/blogs/menasource/from-tripoli-to-tehran-lessons-from-libya-in-us-iran-nuclear-talks/
Wilson Center, “Giving Up on the Bomb: Revisiting Libya’s Decision to Dismantle its Nuclear Program,” 2017. https://www.wilsoncenter.org/blog-post/giving-the-bomb-revisiting-libyas-decision-to-dismantle-its-nuclear-program
Atlantic Council, Danny Citrinowicz, “To preserve ‘Rising Lion’s’ achievements, Israel must support an Iran nuclear deal,” October 6, 2025. https://www.atlanticcouncil.org/blogs/menasource/to-preserve-rising-lions-achievements-israel-must-support-an-iran-nuclear-deal/
