Iran’s nuclear program has advanced centrifuges and clandestine risks since the JCPOA. A new deal must address weaponization, covert facilities, and IAEA access—not just enrichment. Without verification, bombing and moratoria fail.
The Iran Nuclear Program remains the central focus of U.S. foreign policy as conflict persists. While leadership attempts to define objectives, the Iran Nuclear Program has evolved far beyond previous constraints. Negotiating a new deal for the Iran Nuclear Program requires addressing technical leaps. Without oversight, the Iran Nuclear Program could accelerate covertly.
How the Iran Nuclear Program Outpaced the JCPOA
Ever since the United States, alongside Israel, went to war against Iran in late February, U.S. President Donald Trump has struggled to define the objectives of the conflict. His emphasis has shifted from demands for regime change to degrading Iran’s military to securing safe passage through the Strait of Hormuz. Most frequently, he has seemed to emphasize a need to roll back Tehran’s nuclear program.
It is an odd frame for a war that has barely touched Iranian nuclear capabilities, but a fitting one for a leader who withdrew from President Barack Obama’s 2015 nuclear deal with Iran, the Joint Comprehensive Plan of Action, in 2018 and claimed a year ago to have “obliterated” Iran’s program. “The DEAL that we are making with Iran will be FAR BETTER than the JCPOA,” Trump boasted on Truth Social in late April.
But achieving such an agreement will be more difficult than the president seems to realize, in large part because Iran’s nuclear capabilities have advanced since Trump withdrew the United States from the 2015 treaty during his first term. Public reports suggest that ongoing U.S.-Iranian negotiations over Tehran’s program are focused on two elements: the length of a moratorium on its uranium enrichment activities and the fate of its stockpile of highly enriched uranium.
Both are necessary components of any successful future nuclear deal, but they are also insufficient. Over the past seven years, Iran has markedly improved its ability to manufacture and install more powerful enrichment centrifuges, shrinking the time needed to produce the material for a nuclear weapon. And there are now more gaps in international inspectors’ knowledge about the extent of the program.
Technological Sprints within the Iran Nuclear Program
For bragging rights, Trump doesn’t only need a deal that differs from the JCPOA; he needs a dramatically different one. A deal in 2026 must go well beyond addressing enrichment and stockpiles; it must also create new, detailed procedures that allow inspectors to understand Iran’s current capabilities and to prevent the country from making covert progress toward a weapon. Without addressing these concerns, it doesn’t matter how much Washington bombs Iran, or how long an enrichment moratorium lasts, or what happens to the country’s highly enriched uranium. Tehran could emerge from the war closer to a nuclear weapon than it was before.
LEARNING TO SPRINT
Last June, Iran’s enrichment program was severely degraded by U.S. and Israeli strikes in what is now known as the 12-day war. Nuclear scientists were killed, and enrichment capacity at the country’s underground Fordow and Natanz facilities was badly damaged or destroyed. Yet Iran’s nuclear program still poses a greater challenge now than it did in 2015, when the last nuclear treaty was finalized.
The pact capped not only the number of centrifuges Iran could operate but also, and more important, the kinds it could operate or manufacture. And it prevented Iran from carrying out certain research that would advance its centrifuge program. After the United States withdrew from the deal, these provisions were no longer implemented.
In the years that followed, Iran acquired technical knowledge in centrifuge production and operation that cannot be bombed away.
By June 2025, its best centrifuges were roughly six times as efficient as the ones that existed in 2015. Iran also improved its installation speed: the rate at which a country can produce material for nuclear weapons depends not only on how many and what type of centrifuges it has but also on how quickly it can install additional ones into “cascades,” or networks of connected devices that accelerate the concentration process.
In 2015, Iran could install roughly two cascades per month, each of them containing approximately 170 interconnected centrifuges. In 2025, the country demonstrated an ability to install cascades nearly three times faster.
Iran Nuclear Program: The Race to Weaponization
Together, these technological advancements greatly compress the time Tehran needs to produce material for a nuclear weapon, even if it has to rebuild its nuclear infrastructure from scratch. Consider Iran’s ability to produce the material for a nuclear weapon as a race on a track on which the pace is set by the country’s technological prowess and the starting point is set by the severity of the constraints.
From the starting point imposed by the JCPOA, Iran could have covered the distance in about a year, walking at a moderate pace. Supporters of the treaty argued that this lag would give other countries adequate time to detect an attempt at nuclear breakout and respond diplomatically or militarily before Iran reached the finish line.
But between 2018 and 2025, Iran learned to sprint. As a result, even if Iran accepted Trump’s demands for zero enrichment and gave up all its highly enriched uranium—more severe constraints than the 2015 deal imposed—it might still be able to finish the race in less time.
Indeed, even if Iran made those concessions and last year’s 12-day war fully destroyed all the centrifuges at Natanz and Fordow, we estimate that Tehran would need closer to six months than a year to rapidly install more capable centrifuges somewhere else and produce one weapon’s worth of enriched material. The timeline grows even shorter if Iran keeps some of the thousands of kilograms of lower-quality enriched uranium it is known to possess.
Gaps in Monitoring the Iran Nuclear Program
GAPS IN KNOWLEDGE
Iran’s known nuclear sites are under rubble. But there is still great uncertainty about whether it might have other enrichment-related facilities.
The country has a long history of pursuing enrichment clandestinely; Iran’s two declared enrichment facilities at Natanz and Fordow were both built covertly at the turn of the twenty-first century, violating the country’s legal obligation to inform the International Atomic Energy Agency that it was creating such sites.
In early 2025, Iran publicly announced a plan to build a third underground enrichment facility at Isfahan, and satellite imagery released before the 12-day war that June increased concerns that another substantial one might be under construction at Pickaxe Mountain near Natanz. Because the IAEA has not visited either of the sites, their operational status is unknown, just as it is for any other clandestine facilities.
Throughout the course of diplomacy to contain Iran’s nuclear ambitions, efforts have been made to address the possibility that Iran would simply fail to disclose key activities. In 2003, after an Iranian opposition group exposed the existence of the Natanz facility, Iran temporarily agreed to allow IAEA inspectors to implement the Additional Protocol, a legal instrument expressly designed to help them discover covert nuclear activities.
The 2015 nuclear accord, for its part, specifically required Iran to reimplement the Additional Protocol and allow extra oversight of Iranian centrifuge production. By tracing each centrifuge Iran manufactured, inspectors could confirm that all of them had been installed in declared, monitored facilities and not diverted to unknown destinations.
This IAEA centrifuge monitoring program continued after the United States withdrew from the pact, but it collapsed in 2021 when Iran halted cooperation in retaliation for the assassination of the Iranian nuclear scientist Mohsen Fakhrizadeh by Israeli operatives. Since that oversight program ended, Iran has produced tens of thousands of centrifuges.
Most were certainly installed at Fordow and Natanz, but if even a few hundred were diverted to other, clandestine facilities, they could quickly be used to produce weapons-grade material from low-enriched stocks.
Reforming Oversight for the Iran Nuclear Program
The 2015 treaty ultimately focused more on constraining Iran’s ability to enrich uranium for a nuclear weapon than on limiting its ability to assemble actual nuclear arms. That was because restricting enrichment is easier than monitoring and derailing all the activities that subsequently contribute to assembling a weapon.
Enrichment requires specialized industrial facilities, whereas weaponization entails things like computer modeling, conventional explosives testing, and work on warhead design, all of which can be conducted in small facilities that are far more difficult to distinguish from those used for ordinary scientific or industrial research. And for many years, U.S. intelligence assessed that Iran was not undertaking key weaponization activities needed to produce a nuclear device.
Section T of the JCPOA did seek to address the challenge of limiting activities relevant to the design of a nuclear weapon, but the enforcement mechanism was left unclear and the provision went untested.
The nature of Iran’s nuclear challenge, however, has fundamentally shifted since 2015. Building a nuclear weapon requires two things: making the necessary fissile material and turning that material into an actual warhead.
When producing fissile material was the dominant obstacle, focusing less on weaponization was an understandable choice. But Iran’s work to dramatically accelerate its enrichment timeline has changed that calculus. And the concern that Iran could pursue weaponization has grown more concrete: in 2024, U.S. intelligence agencies quietly dropped their long-standing assessment that Iran was not pursuing such activities from their annual report.
Preventing and detecting weaponization activities is difficult because it requires new tools and new authorities, but the world can no longer avoid doing it. Securing the kind of access that would allow monitors to adequately oversee weaponization would be difficult under any circumstances. But now it will be considerably harder, as Iran claims that IAEA inspections helped facilitate espionage that led to the 2025 and 2026 strikes on its sites.
Future Stability and the Iran Nuclear Program
A GOOD AGREEMENT
It is still in the U.S. national interest to find a diplomatic solution to the Iranian nuclear threat. But what worked in 2015 is no longer enough. Removing stockpiles of uranium enriched to 60 percent—what Trump refers to as “nuclear dust”—would eliminate the most pressing concern, and a full suspension of enrichment for five years, as Iran has reportedly proposed, would go beyond certain restrictions in the 2015 accord.
But a nuclear deal focusing only on enrichment and stockpiles can no longer address the problem. A good agreement must now also account for Iran’s improved enrichment technology, its potential for covert enrichment activities, and its ability to convert enriched uranium into deployable weapons.
To ensure that Iran cannot agree to a moratorium on enrichment while busily enriching at clandestine facilities, a new deal must require Tehran to resume its implementation of the Additional Protocol, which was specifically designed to help the IAEA investigate and unmask secret nuclear facilities.
As part of such an accord, agency inspectors must be tasked with making the most complete accounting possible of all the centrifuges Iran has produced since 2021, when the country banned advanced monitoring. The fact that both Iran’s manufacturing facilities and its centrifuges may now lie under rubble will significantly complicate this work.
To limit Iran’s ability to convert uranium into weapons, Tehran must also provide verifiable information about nonnuclear research and military activities that have potential weapons applications and ensure that the IAEA can inspect military sites. Implementing existing provisions in the Additional Protocol and adding ones similar to Section T of the 2015 agreement will help. But these tools must be further expanded to address weaponization activities and be matched with clear stipulations for access.
Iran will resist demands that the IAEA be allowed to inspect military facilities. But any ambiguity about the agency’s right to pursue suspected weaponization activities must be resolved now, not left to future interpretation.
That the Trump administration has made little public mention of the roles that verification and IAEA monitoring must play in a deal with Iran is a very worrying sign. Overall, a lack of advance planning has characterized the United States’ entire war effort: the Trump administration clearly put insufficient work into mapping the conflict’s potential trajectory and its impact on the global economy, Washington’s closest alliances, and U.S. credibility in general. It would be a big mistake if the same lack of attention to detail pervades Washington’s efforts to negotiate a new nuclear deal.
NECESSARY BUT NOT SUFFICIENT
The recent war has made it obvious that not even a perfect nuclear deal can address the threat Iran poses. In truth, the United States’ struggle to define its priorities with respect to Iran long precedes this administration. Both of us participated in past U.S. government efforts to address the Iranian nuclear threat, Swanson as an Iran specialist and Sharp as a technical expert focused on nuclear issues. Sometimes, Washington made Iran’s nuclear program the fulcrum of U.S. policy on Iran. At other times, policymakers deprioritized the nuclear threat and focused on the totality of the risks posed by Iran, and on the regime’s treatment of its people.
The Iran war has taught us that neither approach can work alone. Focusing narrowly on Tehran’s nuclear program cannot resolve Iran’s threat to U.S. security. The most immediate threats that the country poses to U.S. interests are its ability to control ship traffic through the Strait of Hormuz and its missile program, not its nuclear program. But a nuclear-armed Iran would create its own, terrible problem and make these existing threats immeasurably graver.
A nuclear deal is both necessary and insufficient for addressing the broader challenge, one that is only growing more protean and dangerous as time elapses and that cannot be addressed solely by military means.
If Trump wants to strike a new nuclear pact with Iran, he therefore has to make it a good one. A deal that only addresses known enrichment sites and uranium stockpiles and does not attend to the country’s increased enrichment capacity or the possibility of covert activity or weaponization will not prevent Iran from obtaining a nuclear arm. It will only push Iran’s effort further underground—and make a future solution even harder to find.