May 2026
By Matthew Bunn
U.S. President Donald Trump’s April 1 address to the nation on the war on Iran offered a torrent of lies and no solutions.1 Far from freeing the United States from “the specter of nuclear blackmail,” it appears that this war will leave a greatly weakened but embittered and harder-lined Iranian regime that is more determined than ever to build a nuclear deterrent. Although on-again, off-again talks are still underway, as of late April, Iran still retains much of the material and equipment needed to do so. That includes enough weapons-usable highly enriched uranium (HEU) for more than 10 nuclear bombs.

The issue is not just the HEU, however. After decades of effort by thousands of people, Iran has an array of material, equipment, and expertise that, although damaged, cannot be fully bombed away. The war will leave Iran’s government less able to defend itself but the danger of Iranian nuclear weapons will remain.
It is worth remembering how we got to this point. Under the 2015 Joint Comprehensive Plan of Action (JCPOA), better known as the Iran nuclear deal, Tehran agreed to stringent limits on its uranium enrichment, exported 97 percent of its enriched uranium to Russia, and accepted inspections going far beyond usual International Atomic Energy Agency (IAEA) safeguards.2 But Trump pulled out of the deal in his first term, freeing Iran to produce HEU, develop more advanced centrifuges, and abandon the deal’s inspection provisions.
If Trump had not pulled out of the JCPOA, Iran’s stockpile of HEU—its shortest path to the bomb—would not exist. Trump’s claim that if he had not pulled out, Iran would now have nuclear weapons is simply a lie. (If, however, the United States and its partners had wasted the time the deal offered and no further accords had been reached, the sunset clauses on the limits Iran had accepted would be starting to kick in.) Moreover, before this war, Iran offered to blend down its HEU so it could not be used in weapons, pause all enrichment for several years, and not stockpile enriched uranium—an offer Trump rejected, without exploring whether it could be improved, in favor of war.3
To be fair, there is no doubt that the military attacks during Trump’s second term have been a major setback for Iran’s nuclear program. Iran’s major non-reactor nuclear facilities were seriously damaged in the airstrikes Israel and the United States carried out in June 2025; those attacks also killed several prominent scientists and engineers associated with Iran’s nuclear program. None of its uranium enrichment or conversion facilities appear to have been operable after those strikes.4 The current war, which began February 28, appears to have done some modest additional damage to Iran’s nuclear efforts, including hitting the entrance to the Natanz enrichment facility with bunker-buster bombs, further ensuring it could not readily be returned to operation.
Nevertheless, although this war will have lasting and far-reaching effects on everything from global geopolitics to oil prices to the U.S. reputation in the world, there is little evidence it will prevent a future Iranian nuclear weapon. Iran’s nuclear program is far from “obliterated.” Indeed, by dramatically increasing Iran’s resolve to acquire a nuclear deterrent and leaving key material and some equipment and expertise in place, I believe that the war’s net effect will be to increase, rather than decrease, the probability that Iran will have nuclear weapons a decade from now.
Key issues in assessing this judgment include the fate of Iran’s HEU; its other stocks of enriched uranium; its centrifuges and components to make them; and its weaponization-related equipment and expertise. They also include Iran’s incentives to pursue nuclear weapons; the ability of intelligence agencies to detect a future move toward nuclear weapons in time to stop it; and the need, ultimately, to find a new path to agreed restraints and effective international monitoring.
The Dangers of Iran’s Surviving HEU
The most immediate concern is the roughly 441 kilograms of HEU enriched to 60 percent U-235 that Iran had accumulated before the summer 2025 strikes.5 This material was stored in the form of uranium hexafluoride (UF6), in canisters similar to scuba tanks. Most, perhaps all, of this material appears to have survived last summer’s strikes and this year’s war.6 This is the material that Iran could most readily use to produce nuclear weapons, and it might pose broader hazards as well.
First, Iran might choose to enrich this material to 90 percent or more U-235 to use it in the weapon design that Iran was working on in the secret nuclear weapons program it largely stopped in late 2003.7 With just 100 of the advanced centrifuges it has developed since Trump’s pullout from the nuclear deal in his first term, Iran could further enrich enough material for a bomb in just a few weeks. If Iran still has access to all of the 60 percent enriched HEU stock, it would be enough, once further enriched, for more than 10 nuclear weapons.
Second, even without further enrichment, the 60 percent enriched material, once converted to metal, could be used in a bomb. It could be put in a weapon of the same design, exploding with less power, or it could be used in a modified design using more material and explosives.8 Indeed, this material could be used to make several simple “gun-type” nuclear weapons comparable to the Hiroshima bomb, which could be made quickly, although Iran would likely have to rely on covert means of delivery as the bombs would be too heavy for its missiles.
Beyond Iran using this material itself, there are other concerns. Nobody knows who might take possession of it in the resulting chaos if Iran’s government collapses. Even without a government collapse, the current crisis might lead some lower-level people managing the HEU to decide they needed to get out of the country; they might see selling this material as their golden ticket, although transporting the material in its current form would be a challenge. If the material falls into the hands of terrorists, multiple studies by the U.S. and other governments have warned that a sophisticated extremist group with HEU might well be able to make a crude bomb. As with the Iranian government, however, they would first have to convert it to metal, which would be an additional challenge.9
Where is Iran’s HEU Now?
A critical question is what has happened to this dangerous material as a result of the U.S. and Israeli attacks. Logically, there are three possibilities: It was in canisters that were destroyed in the June 2025 bombings and is no longer available because it is scattered over the walls, floors, and rubble; it was in canisters at sites that were bombed but the canisters remained intact and could be recovered by digging down into the facilities; or, it is in canisters that were at facilities that were not destroyed and could be recovered.
Unfortunately, it appears that most of Iran’s HEU is in this third category, although possibly blocked by rubble at the entrances that would have to be cleared to access it. The IAEA has noted that before the 2025 strikes, much of the HEU, as well as Iran’s 20 percent enriched U-235, was stored in deep underground tunnels near the nuclear facility outside the city of Isfahan.10 In particular, IAEA Director-General Rafael Mariano Grossi has reported that more than 200 kilograms of it was stored there before the strikes. The IAEA has not been permitted to visit since then.11
Some reports suggest that other portions of Iran’s HEU were stored in a deep underground facility near Natanz known as Pickaxe Mountain, and in the enrichment site at Fordow, one of the locations bombed in 2025.12 It is at least possible that Iran removed the HEU from some or all of these facilities before those strikes, as vehicles were seen at these facilities in the days before military action began.
In late March 2026, a satellite photograph surfaced that was taken June 9, 2025, just days before last year’s strikes began, showing a flat-bed truck outside the entrance to the Isfahan tunnel complex carrying more than a dozen 1.1-meter diameter containers. The truck was accompanied by another truck with a mobile crane, perhaps for lifting the containers, and vehicles that might be security vehicles. No definitive judgments can be made from this single photograph: it does not reveal what was in the containers, or even whether the truck was going into the tunnel complex or backing out of it. One plausible explanation is that Iran moved even more HEU into the Isfahan tunnel complex as protection from the attacks it expected were coming. The containers could be HEU canisters with overpacks for safe transport.13
The tunnels at Isfahan are significantly deeper than the underground Fordow enrichment facility that was seriously damaged by U.S. “Massive Ordnance Penetrator” bombs in the 2025 strikes, and they have not been destroyed. Gen. Dan Caine, the chairman of the U.S. Joint Chiefs of Staff, has reportedly acknowledged that the Isfahan tunnels are too deep to destroy with current bunker-buster bombs.14 The tunnels at Pickaxe Mountain are also deep and the mountain is granite, making those tunnels at least as challenging a target.
Although the tunnel complex near Isfahan has not been destroyed, the entrance was bombed, raising the question of how difficult it would be to get at the containers in the tunnels. Trump has asserted that “it would take months” to dig down to get the “nuclear dust” in these tunnels; in one interview, he said the HEU was “so far underground, I don’t care about that.”15
Yet a wide range of public evidence suggests Trump is wrong about how hard this material would be to reach. First, in bilateral talks before the ongoing war, and in the Islamabad talks during the ceasefire, the Trump administration demanded that Iran export this material, and Iran, before the war, offered to blend it down—implying that both sides thought Iran could get access to it. Second, substantial press reporting indicates that planning for a possible operation to seize or destroy this HEU includes construction equipment to excavate the rubble, suggesting that this work could be done fairly quickly.16 Third, the IAEA has made clear that it sees this material as a major continuing concern; it has observed “regular vehicular activity” around the tunnel entrances since the bombings last summer, although Grossi has said that activity did not include major digging.17 Finally, there are some reports that the material is already accessible.18
What is Next for Iran’s HEU?
If airstrikes alone will not solve the problem of the HEU in these deep tunnels, there are a few other options. One is surveillance and deterrence. Trump seems to be leaning toward simply leaving the HEU where it is and watching to make sure Iran does not remove it from the tunnels. As he put it in his April 1 address, the United States has the HEU “under intense satellite surveillance.… If we see them make a move, even a move for it, [we] will hit them with missiles very hard again.”19 Leaving more than 10 bombs’ worth of HEU in Iran, however, while relying only on a return to war to prevent its use, is hardly an effective long-term solution.
Another route the Trump administration is trying to pursue is negotiating a deal with Tehran to remove the material. In the JCPOA, Iran agreed to export nearly all of its enriched uranium to Russia. In the talks before the ongoing war, and apparently again in the Islamabad talks, the Trump team demanded that Iran export the HEU to the United States; Iran rejected that idea but offered to blend it down to lower levels of enrichment in Iran. As of mid-April 2026, there were no signs that Iran was willing to accept U.S. demands for exporting this material.
A deal, if one could be reached, would be a far more practical and less dangerous approach than any military option. Over the years, HEU has been removed from many cooperative countries. One early example was Project Sapphire, in which U.S. teams worked with Kazakhstan in 1994 to fly 1,280 pounds (580 kilograms) of HEU to safe storage in Tennessee.20 Programs such as the Global Threat Reduction Initiative and its successor, Material Management and Minimization, have removed tons of plutonium and HEU from scores of sites around the world. Indeed, more than half of all countries that once had HEU or plutonium separated from spent fuel on their soil have shipped it back to its country of origin, ensuring that terrorists will not be able to obtain weapons-usable nuclear material from the emptied locations.
Even a cooperative approach would not be easy. Whatever rubble may be blocking the entrances to the site or sites where the Iranian HEU now exists would have to be cleared and the HEU canisters would have to be assessed, measured, and carefully packaged. Then, they would have to be transported out of the tunnels, trucked to an available airstrip, and flown to a country willing to accept the HEU.
Without Iran’s cooperation, the remaining option for eliminating the HEU would be sending in a team of U.S. or Israeli soldiers and experts to deal with the material. U.S. special operations forces have long trained with federal scientists and experts to disable or secure adversaries’ nuclear weapons and material,21 but a military operation would not be easy. Removing the material would require taking all the steps of the cooperative approach, except it would be in the middle of a war zone. There are probably dozens of containers, collectively weighing tons, in multiple locations deep inside Iran—a country as big as France, Germany, and Spain combined. Troops would probably need to collect the material from several facilities (possibly including digging through rubble at the entrances), secure an airstrip near each, truck or helicopter the equipment and material to and from the strip, and defend against attacks on the preparations and shipments. Flying everything out of Iran in the face of remaining air defenses would be another hazardous operation.

U.S. and Israeli air power might be able to devastate Iranian forces near the HEU sites, clearing the way for the deployment of U.S. troops. Trump has said he would only authorize a ground operation if Iran was “so decimated that they wouldn’t be able to fight on the ground level.”22 However, there may well be extensive booby traps and other defenses in the tunnels (and possibly decoy canisters that would have to be distinguished from the real ones), and Iranian missiles and drones could threaten the troops on the surface. Mark Esper, a defense secretary in Trump’s first term, has warned that sending ground troops into Iran to address the HEU stockpile would take a large force and be “very perilous.”23
A simpler but messier option than removal would be to blow up the containers, with explosives attached to each one. The UF6 would end up on the walls, floors, and rubble in the tunnels, making it difficult to ever recover and use. The tunnels would then be contaminated and unusable, and the U.S. team would need to be careful about its own safety.
Blending the material to low-enriched uranium (LEU) so it could not support an explosive nuclear chain reaction is another option. Doing that on-site would also be difficult, requiring the delivery of equipment and tons of uranium for blending into an active war zone. The National Nuclear Security Administration has developed mobile equipment for packaging HEU in the past, although it has never been used to downblend hundreds of kilograms of UF6, let alone doing that in a war zone.24 The result of such a blend-down would be many tons of LEU. Enrichment is a process that speeds up as it goes, so that when uranium has been enriched to 5 percent, two-thirds of the work of enriching to 90 percent is already done. That is why the JCPOA limited Iran to only 300 kilograms of LEU at a time. Leaving many times that amount in Iran would pose a serious risk that it could be re-enriched in the future.
Could Washington know for sure that it had secured all the HEU, whether by the cooperative or non-cooperative options? Before the June 2025 strikes, the nuclear material was all monitored by the IAEA. The agency knows how many canisters there were, and the serial numbers of each one, but no international inspectors have seen this material since then. What if U.S. forces found 70 percent of those canisters? Would they be confident the rest were deeply buried under rubble at such sites as Natanz and Fordow, and not held in some secret stockpile?
Iran’s Other Uranium
HEU is not the only issue. Before the June 2025 strikes, Iran also had 181 kilograms of uranium enriched to 20 percent U-235, more than six metric tons of 5 percent enriched LEU, more than two metric tons of 2 percent enriched LEU, and likely a large amount of uranium in UF6 that was not yet enriched.25
Material such as 5 percent LEU cannot be used directly in a nuclear bomb, but as just noted, it would provide a two-thirds head start on enriching all the way to 90 percent. If further enriched, the stock of 20 percent enriched material would be enough for one to two nuclear weapons, whereas the stock of 5 percent would be enough for roughly a dozen.
What has happened to this material? The IAEA reports that the 20 percent enriched uranium also was stored in the deep tunnels at Isfahan, and thus it, too, is probably available to Iran, with some effort. The 5 percent, 2 percent, and unenriched material was stored in huge 30B containers, weighing nearly three metric tons when full. Many of these containers are thought to have been at the above-ground nuclear facility at Isfahan, where the UF6 is produced from uranium oxide, but some may have been at the bombed enrichment plants at Fordow and Natanz, and may be in either intact or ruptured containers buried in the rubble.26 Publicly at least, little is known about how much of this material may have survived; U.S. and Israeli intelligence agencies presumably know more, but this stock again raises the question of how confident the U.S. and Israeli governments could be that there were no secret stockpiles.
Iran’s Centrifuges and Centrifuge Parts
The 2025 U.S. and Israeli attacks severely damaged Iran’s main enrichment sites at Natanz and Fordow. It is unlikely that any of the centrifuges at either site are still operable. Other strikes hit known facilities for assembling centrifuges and making key parts.
Nevertheless, there are serious questions about what centrifuge capabilities still remain. As Grossi has noted, at the time of the JCPOA, Iran was primarily relying on “a very primitive type of centrifuge.” Since Trump’s pullout from the nuclear deal, there have been no restraints on Iran’s development of more advanced machines, and it “has the most sophisticated, fast, and efficient” centrifuges, meaning that the enrichment needed to produce bomb material could be done faster, by fewer centrifuges, in an easier-to-hide facility.27
One innovation of the JCPOA was that it established a system of monitoring Iran’s production of centrifuges, but that ended after Trump pulled out of the accord; Iran has been producing centrifuges and centrifuge parts for years without any monitoring. It would be surprising if Iran had not dispersed centrifuges and the parts for them to a wide range of locations; Grossi has warned that there may be “dozens of workshops” for this purpose.28 It is hard to know whether U.S. and Israeli intelligence have been able to find and target all those places.
Pickaxe Mountain is a particular concern. Iran has described that facility as a plant for assembling centrifuges, yet reportedly it was not yet operational when the 2025 attacks began. There may well be stocks of centrifuges and the parts to assemble more there, and like the tunnel complex at Isfahan, it is too deep to be reached by U.S. bunker-buster bombs. More worrisome, the complex there is too big for just assembling centrifuges—big enough to assemble them, set up cascades to enrich uranium, have facilities to convert the uranium to metal, and then turn the metal into nuclear weapons.
Weaponization Equipment and Expertise
From the late 1990s until fall 2003, Iran had a serious covert program to design and build nuclear weapons and carry out a nuclear test. The country’s supreme leader, Ayatollah Ali Khamenei, ordered a halt to the visible weaponization parts of that program in late 2003, while enrichment efforts continued apace. Part of the dispute between Iran and the IAEA has had to do with what happened to materials and equipment from that past program.29 Since Trump pulled out of the JCPOA, there have been new or renewed activities in Iran that Grossi has described as “concerning.”30 Meanwhile, U.S. intelligence agencies dropped from their annual threat assessment that Iran had not resumed its weaponization effort.
During the 2025 strikes, several locations that the Israeli Defense Forces (IDF) asserted were associated with potential weaponization activities were destroyed, most likely along with some equipment related to weaponization. The necessary equipment, however, is not especially unique or difficult to rebuild.
Iran’s deeply buried facilities, including the tunnels at Isfahan and the facility at Pickaxe Mountain, remain an ongoing concern. They are largely immune to air attack, and one effect of the 2025 strikes was to educate Iran about how deep they had to dig to protect their facilities. It is difficult to know what equipment and materials may be stored there—or what secret locations intelligence agencies have not yet managed to locate.
The loss of the people who have been killed by Israeli strikes might inhibit an Iranian effort to rebuild a weapons program even more than the loss of equipment and facilities. After several assassinations in previous years, including Mohsen Fakhrizadeh, leader of the past nuclear weapon program, the IDF asserted that it killed some 20 weapon scientists and engineers in the 2025 strikes and has announced additional deaths during the 2026 war.31
Never before has a state undertaken such a large-scale killing campaign against the people participating in a nuclear program. The loss of these figures is undoubtedly a setback for any effort Iran might undertake to pursue nuclear weapons in the future. But how large a setback? The overall Iranian nuclear program has involved thousands of people, and the functions most related to weaponization likely had scores to hundreds of participants. Clearly, some people are more important to an overall program than others, so it may be that the people who have been killed were very central to the effort. Without detailed intelligence information, the overall impact is difficult to judge. Yet it seems very likely that with time, other experts will be able and willing to take the place of those who have been killed, should Iran decide to pursue nuclear weapons.
Nuclear-Weapon Incentives
Such a decision appears to be a real possibility. After this war, Iran is likely to have stronger incentives to build nuclear weapons than ever before. It has been attacked twice in the past year by powers with overwhelmingly greater conventional military power. The tools it relied on before this to deter attack—the regional groups it armed and supported, its missile fleet, and its nuclear latency—have utterly failed and are far weaker than they once were. Hard-line members of parliament have already submitted legislation requiring Iran to withdraw from the nuclear Nonproliferation Treaty (NPT), although it is not clear when that could be voted on, as parliament has not been gathering during the war. The NPT “has had no benefit for us,” according to Ebrahim Rezaei, a spokesman for the Iranian parliament’s National Security Commission.32
Khamenei, the supreme leader who ordered a halt to the weaponization program in late 2003 and wrote a fatwa prohibiting nuclear weapons, is dead. Mojtaba Khamenei, his son and successor, is reportedly even harder-line and closer to the Islamic Revolutionary Guard Corps (IRGC) than his father.33 The opening attacks of the 2026 war reportedly killed his father, mother, wife, teenage son, sister, and brother-in-law and seriously injured him.34 That hardly seems likely to incline him toward a softer line toward the countries that attacked. He reportedly opposed his father’s fatwa against nuclear weapons and could readily reverse it. Below Khamenei, it appears that hard-liners in the IRGC, many of whom have long believed that Iran should build nuclear weapons, are increasingly in charge.
In short, it appears that although the June 2025 attacks and the 2026 war have reduced Iran’s capability to develop and build nuclear weapons, these attacks have built incentives for Iran to do so that are greater than ever before and put people in charge who are likely to be inclined to move in that direction.
The major problem Iranian bomb advocates face is the risk that such an effort would again be attacked before it could be completed. The U.S. and Israeli attacks of June 2025 and the 2026 war have revealed just how deeply Iran’s nuclear program was penetrated by Israeli and Western intelligence agencies, who were able to find and target a range of secret facilities and nuclear program participants. Iran has arrested hundreds of people in its effort to eliminate the spies, but clearly, any effort to launch a renewed nuclear weapon program at secret facilities would run a serious risk of being discovered and attacked.
Relying on Intelligence Success
The impressive military successes of the June 2025 strikes and the current war have been based on even more remarkable intelligence successes, which uncovered what Iran was doing in its nuclear program, where it was taking place, and, in many cases, who was doing it, down to the level of which bedroom particular people were sleeping in.
At the same time, the attacks have meant an end to any international inspection of Iran’s key nuclear facilities. Although Iran permitted safeguards inspectors to return to the nuclear power reactor at Bushehr and the Tehran Research Reactor, no one from the IAEA has seen any of Iran’s enriched uranium, centrifuges, or other key equipment since before the June 2025 attacks. The United States, Israel, and other countries concerned about Iran’s nuclear program have had to rely entirely on the continuing success of intelligence agencies in finding every secret nuclear effort.
So far, that success has been impressive, dating back many years, from the discovery of the secret Fordow uranium enrichment plant before the first centrifuge was installed in the 2000s to Israel’s theft of the vast archive of nuclear weapons documents from Iran. But how long will that success last? Some human and electronic sources have probably been destroyed or compromised during the recent attacks, and Iran’s government may have succeeded in uncovering others in its intensified counter-spy efforts.
The fact is that a nuclear weapons program need not be large. An enrichment plant capable of producing enough HEU for a bomb every year might be smaller than a typical supermarket and use a comparable amount of power. The facility for converting that uranium to metal and machining the metal into bomb components might be smaller still. The facility for assembling the nuclear weapons and fitting them to delivery vehicles would not need to be large either. Will intelligence agencies always be able to find these in time?
Consider, for example, the facility that Iran was building for manufacturing nuclear weapons decades ago, when its nuclear weapons program was at full throttle. When the program halted in 2003, Iran had dug out tunnels under a mountain for this facility and purchased the necessary equipment, although it was not yet installed. Neither U.S. nor Israeli intelligence had identified the construction as a nuclear facility. (To be fair, with no nuclear-related equipment installed, that would have been hard to do.)35
The Need for a Deal
At this writing, the Islamabad talks seem to have left a vast gap between American and Iranian proposals. There appears to be a substantial chance that Trump will end the war against Iran with the HEU still in Iran, no agreed limits on Iranian enrichment, no international monitoring in place, and a harder-line regime in charge in Iran. Trump himself has raised the possibility that Iran would renew its nuclear weapons program, forcing a future U.S. president to attack Iran again.36 Any negotiated agreement that led to that result would be justly described as a terrible deal, although Trump is trying to pitch this result as a victory.
Ultimately, Iran’s nuclear knowledge cannot be bombed away. With time and motivation, destroyed facilities can be rebuilt. Diplomacy provides the only long-term approach that can potentially provide lasting, stringent limits, effective international monitoring, and reduce the sense of threat that is the best argument for Iran’s nuclear weapon advocates. Provisions to accomplish those objectives were central to the JCPOA. As Grossi put it:
“[O]nce the military effort comes to an end, we will still inherit a number of major issues that have been at the center of all of this… a lot still has survived. They have the capabilities, they have the knowledge, they have the industrial ability to [pursue nuclear weapons]. This is why we need to go back to a negotiating table.”37
Trump’s team is continuing to seek a deal with Iran on its nuclear program, the Strait of Hormuz, and other matters. But the president seems to believe that making extreme threats and sending negotiators with no nuclear expertise, verification expertise, Iran expertise, or sanctions expertise will be sufficient to reach an effective deal. That stands in stark contrast to the deep expertise the Obama administration’s negotiation team brought to the table—in talks that stretched over years, not hours.
Insisting, as the Trump team reportedly is, on Iran agreeing to zero uranium enrichment forever is neither necessary nor wise. A deal for zero would not buy very much extra security, as Iran could move from zero quickly, having demonstrated in the past that it can install 1,000 centrifuges—enough to launch a bomb program—in a month. Moreover, Iran’s government has consistently rejected an enrichment ban for decades, framing it as colonial powers seeking to take away Iran’s God-given right to technological development. Similarly, insisting that Iran export its HEU to the United States—the country that has just been attacking it—rather than to its uneasy partner Russia, is almost certainly non-negotiable. It is precisely because Russia might be willing to aid Iran if the United States reneged on a deal that a Russian option would have some small chance of being negotiable.
With Trump having pulled out of the JCPOA and twice launched military strikes while talks on new deals were underway, Iran now has very little reason to believe any promises that the United States makes, and a strong incentive to preserve its nuclear options. Nor does Washington have much reason to believe Iranian promises, despite the fact that Iran, unlike the United States, complied with the 2015 nuclear deal. The 2025 strikes and the 2026 war, although setting back Iran’s nuclear capabilities, may well have ruined the chances for negotiating an effective nuclear deal for many years to come.
ENDNOTES
1. Donald Trump, “Address to the Nation on Military Operations in Iran,” The American Presidency Project, University of California, Santa Barbara, April 1, 2026.
2. Gary Samore, ed., “The Iran Nuclear Deal: A Definitive Guide,” Harvard Kennedy School Belfer Center for Science and International Affairs, November 2017.
3. Kelsey Davenport, “U.S. Negotiators Were Ill-Prepared for Serious Nuclear Negotiations with Iran,” Arms Control Association, March 15, 2026.
4. David Albright, Sarah Burkhard, Spencer Faragasso, et al., “Comprehensive Updated Assessment of Iranian Nuclear Sites Five Months After the 12Day War,” Institute for Science and International Security, November 21, 2025.
5. For the quantity, see International Atomic Energy Agency, “NPT Safeguards Agreement with the Islamic Republic of Iran,” GOV/2026/8, February 27, 2026, p. 7.
6. Julien E. Barnes, Tyler Pager, Christian Triebert, Eric Schmitt, and Ronen Bergman, “Iran Could Retrieve Uranium at Site U.S. Bombed Last Year, Officials Say,” The New York Times, March 7, 2026.
7. Aaron Arnold, Matthew Bunn, Caitlin Chase, Steven E. Miller, Rolf Mowatt-Larssen, and William H. Tobey, “The Iran Nuclear Archive: Impressions and Implications,” Harvard Kennedy School Belfer Center for Science and International Affairs, April 2019.
8. Matt Kaplan, “Implosion Nuclear Weapons with 60%-Enriched Uranium,” Science & Global Security, Vol. 33, No. 1-3 (2025), pp. 89-101.
9. For an unclassified example, see Chapter 5, “The Non-State Adversary,” in Nuclear Proliferation and Safeguards, U.S. Congress Office of Technology Assessment, Washington, D.C., OTA, 1977.
10. IAEA, GOV/2026/8, p. 4.
11. Forrest Crellin, “Much of Iran’s near-bomb-grade uranium likely to be in Isfahan, IAEA’s Grossi says,” Reuters, March 9, 2026.
12. Julian Borger, “US weighs sending forces into Iran to secure nuclear stockpile, reports say,” The Guardian, March 19, 2026.
13. See, for example, François Diaz-Maurin, “Analysis: Iran likely transferred highly enriched uranium to Isfahan before the June strikes,” Bulletin of the Atomic Scientists, March 29. 2026.
14. Natasha Bertrand and Zachary Cohen, “US did not use bunker-buster bombs on one of Iran’s nuclear sites, top general tells lawmakers, citing depth of the target,” CNN, June 28, 2025.
15. Trump April 1 address transcript; Steve Holland, “US to leave Iran ‘pretty quickly’ and return if needed, Trump tells Reuters,” Reuters, April 1, 2026.
16. Ellen Nakashima, John Hudson, Alex Horton, and Karen DeYoung, “Risky commando plan to seize Iran’s uranium came at Trump’s request,” Washington Post, April 1, 2026.
17. The vehicular activity is mentioned in IAEA, “NPT Safeguards Agreement with the Islamic Republic of Iran,” p. 4. Grossi mentions the lack of evidence of efforts to recover and remove the material in Crellin, “Much of Iran’s Near-Bomb-Grade Uranium Likely to Be at Isfahan.”
18. Barnes, et al., “Iran Could Retrieve Uranium.”
19. Trump April 1 address transcript.
20. Togzhan Kassenova, “Project Sapphire: How to Keep 600 Kilograms of Kazakh Highly Enriched Uranium Safe,” War on the Rocks, April 1, 2022.
21. See, for example, Walter T. Ham IV, “Highly specialized US Army teams train to disable any potential enemies’ nuclear capabilities,” U.S. Army, January 25, 2024; Joseph Trevithick, “Loose Nukes In Iran Is A Scenario U.S. Special Operators Have Been Training For,” The War Zone, June 18, 2025.
22. Borger, “U.S. weighs sending forces.”
23. Ellen Mitchell, “Esper: Sending special forces to find enriched uranium in Iran would be ‘very perilous’ mission,” The Hill, March 10, 2026.
24. “The Mobile Packaging Program: An origin story,” National Nuclear Security Administration, November 16, 2022.
25. IAEA, GOV/2026/8, p. 7.
26. Interviews.
27. “Transcript: International Atomic Energy Agency Director-General Rafael Grossi on ‘Face the Nation with Margaret Brennan,’” CBS News, March 22, 2026.
28. Ibid.
29. Arnold et al., “The Iran Nuclear Archive.” For a useful summary of the IAEA’s investigations of these issues, see Kelsey Davenport, “IAEA Investigations of Iran’s Nuclear Activities,” Arms Control Association, February 2025.
30. Grossi on Face the Nation.
31. “Significance of the Targeted Nuclear Scientists in the 12-Day War,” Institute for Science and International Security, July 15, 2025.
32. Maziar Motamedi, “As war rages, Iranian politicians push for exit from nuclear weapons treaty,” Al Jazeera, March 28, 2026.
33. See, for example, Karim Sadjapour, “The Iranian Regime Doubles Down,” The Atlantic, March 10, 2026; Patrick Clawson and Farzin Nadimi, “What Kind of Supreme Leader Would Mojtaba Khamenei Be?” Washington Institute for Near East Policy, March 5, 2026.
34. Details of the deaths in that initial strike have not been made public, and accounts vary. This one is based on an interview with Iran’s Ambassador to Cyprus. See Helena Smith, “Mojtaba Khamenei was hurt in strike that killed his father, Iran’s Cyprus ambassador confirms,” The Guardian, March 11, 2026.
35. Arnold et al., “The Iran Nuclear Archive.”
36. David Sanger, “Trump says he halted the nuclear threat from Iran. Evidence suggests otherwise.” The New York Times, March 31, 2026.
37. Grossi on Face the Nation.
Matthew Bunn is a professor of practice at the Harvard Kennedy School and faculty lead for the school’s “Managing the Atom” project.