Reconstructing Iran’s Nuclear Baseline

Editor’s Note: As the United States and Iran renew negotiations regarding what type of nuclear program, if any, Iran will maintain in the future, one of the most difficult issues is the status of the program’s different dimensions. As my Center for Strategic and International Studies colleague Bailey Schiff argues, years without international inspectors, military strikes, and other momentous changes raise numerous questions about the location, scope, and scale of Iran’s existing program. She argues that any deal must ensure inspections resume with full Iranian cooperation.

Daniel Byman

***

The new U.S.-Iran memorandum of understanding gives negotiators 60 days to reach a nuclear deal. But after six years of restricted inspections, U.S. and Israeli attacks on Iran’s nuclear program in 2025 and 2026, and with Tehran’s continuing incentive to maintain ambiguity, the United States does not have a baseline of Iran’s nuclear materials, facilities, and production capacity as it enters negotiations. Reestablishing a baseline will be critical to detecting clandestine Iranian efforts, accounting for nuclear material moved or diverted around the strikes, and designing an effective International Atomic Energy Agency (IAEA) monitoring regime around Iran’s nuclear capabilities today.

Before the 60-day deadline expires, the U.S. delegation should secure three verification provisions: a declaration of activities conducted during monitoring lapses, a comprehensive accounting of remaining materials, and a set of novel inspection authorities. Securing these would strengthen the verifiability of any deal that emerges, deny Iran the leverage that uncertainty provides, and prevent the transparency gaps that hollowed out previous agreements.

The JCPOA’s Legacy

The 2015 Joint Comprehensive Plan of Action (JCPOA) established one of the most extensive verification regimes the IAEA had ever administered, and its monitoring architecture offers lessons for any 2026 agreement, in both its successes and its shortcomings. The JCPOA mandated that Tehran reduce the number of operating centrifuges to 5,060 IR-1s, cap enrichment at 3.67 percent, and cut its enriched uranium stockpile to 300 kilograms (kg). Together, these restrictions allowed the United States to maintain Iran’s breakout time to develop enough fissile material for a nuclear weapon at a year or more.

The agreement also provided authorities beyond those of a comprehensive safeguards agreement, which verifies the peaceful use of nuclear material, or an additional protocol, which grants broader access. JCPOA-specific measures included continuous surveillance of centrifuge component production, real-time enrichment monitoring, a procurement channel requiring advance approval for dual-use imports, and daily inspector access at enrichment facilities. These provisions represent the floor for monitoring that any 2026 agreement will need.

Despite Iran’s nominal compliance with monitoring at declared sites, the JCPOA struggled to verify activity beyond them. Notably, Tehran contested the Section T authority to inspect weaponization-related sites, refusing access at military bases because Iranian officials claimed it would “constitute a national security breach.” Further, critics argued that the IAEA’s closure of its investigation into the possible military dimensions of Iran’s program, while in accordance with the agreed-upon time frame, left key questions unresolved. One arrangement that drew criticism allowed Iran to collect its own environmental samples at Parchin, a site linked to weaponization work before 2004. Former IAEA Deputy Director General Olli Heinonen argued that such procedures and the inability to interview key personnel undermined definitive conclusions about Iran’s program. Mossad’s January 2018 Nuclear Archive operation reaffirmed these fears when it uncovered documents on undeclared high-explosive and detonator work under Project Amad and ultimately led inspectors to a Turquzabad warehouse with man-made natural uranium particles that Iran could not credibly explain. The U.S. withdrew from the JCPOA in May 2018, with President Trump citing concerns about sunset clauses under which material constraints would expire within 15 years, insufficient “mechanism[s] for inspections and verification,” and the exclusion of Iran’s proxy network from the deal. Iran began exceeding JCPOA caps on enrichment the following year.

Still, the JCPOA negotiators inherited a verification regime that does not exist for the 2026 delegation. In 2015, talks took place alongside an active IAEA presence on the ground, maintaining continuity of knowledge at declared facilities under Tehran’s Comprehensive Safeguards Agreement. Today, Iran operates without IAEA surveillance, denies inspectors access to struck facilities, and remains in noncompliance with the Comprehensive Safeguards Agreement. Consequently, determining the current state of Iran’s nuclear program will require resolving open questions on activity during monitoring lapses, and sustaining this baseline knowledge will demand more intrusive IAEA oversight than the JCPOA achieved.

How Iran’s Baseline Broke Down

Any agreement must now contend with two distinct knowledge gaps that will require different solutions and legal authorities to reconstruct.

Unmonitored Activity Since 2021

The first gap concerns centrifuge production and undeclared enrichment capacity. By February 2021, Iran ended its provisional Additional Protocol implementation and JCPOA-specific transparency measures. Over the following months, Tehran denied the IAEA access to centrifuge manufacturing facilities (in Karaj, Natanz, and Isfahan), enrichment plants (Natanz and Fordow), and heavy-water and uranium ore-concentrate sites. By June 2022, it had removed all remaining JCPOA surveillance equipment. The agency has since stated that it “lost continuity of knowledge,” the unbroken inspection and surveillance record that allows it to verify that a state’s declarations are complete. It describes knowledge gaps across centrifuge production and inventory, rotor and bellows manufacturing, heavy water, and uranium ore concentrate as “not possible to restore.”

The production of gas centrifuges sits at the center of this uncertainty. Asked in 2023 about the possibility of Iran diverting centrifuges since restricting IAEA access, Director General Rafael Grossi said plainly, “We don’t know.” Iran Watch estimates that if Iran produced 10 percent more centrifuges than it installed at declared sites between February 2021 and May 2025, it would possess more than 1,300 unaccounted advanced centrifuges, enough to stock a small covert enrichment facility.

The airstrikes in 2025 and 2026 increased the potential importance of this pathway. Power loss, bomb shockwaves, and structural collapse likely destroyed operating centrifuges at Natanz and Fordow, leaving any surviving machines outside declared facilities as the most viable and least detectable pathway for reconstituting an enrichment program. In theory, a facility with enough IR-6 centrifuges could enrich Iran’s remaining 60 percent stockpile to weapons-grade in under a month. Still, a hasty poststrike reconstitution effort would likely take longer due to the challenge of building cascade infrastructure, feed handling, and power and cooling under wartime conditions.

The more difficult question is not whether Iran could improvise such a facility today, but whether one already exists. This is why Pickaxe Mountain, a suspected enrichment plant under construction since 2020 that sits deeper than Fordow, has become a focal point. At a March 9, 2026, press conference, Trump alluded to a site where Iran was resuming work, noting it was “protected by granite,” a characterization consistent with publicly available imagery of Pickaxe Mountain. Though the administration has not clarified which site he meant, satellite analysis shows ongoing construction and fortification of the mountain’s entrances and perimeters. The IAEA hasn’t inspected the site, underscoring why any future agreement must secure complementary access to suspected enrichment sites.

Even with expanded access, reconstructing what happened during monitoring gaps would be a massive undertaking that would require full Iranian cooperation and a verification budget far greater than the pre-JCPOA level. It would first require an Iranian declaration, followed by IAEA access to conduct damage assessments and item-by-item counting and tagging of rotors and bellows, as well as the machines that produce them. Even then, the effort would likely remain incomplete, in part due to the loss of institutional knowledge on the Iranian side. Strikes likely degraded some records at nuclear sites, key scientists have been assassinated before and during both military operations, and Iran likely compartmentalized its nuclear program to prevent leaks, restricting access to a small cadre of specialists. As a result, even a fully cooperative government may be unable to reconstruct a complete account of its own activity.

Declared Materials Since 2025 Strikes

The second gap concerns the location, condition, and quantity of Iran’s declared stockpile since the June 2025 U.S. and Israeli strikes. Establishing what Iran moved before the strikes, and what those strikes damaged or buried, will shape both the framework for resolving the stockpile question and the quality of any negotiations around it. As of the IAEA’s prestrike estimates in June 2025, Iran possessed roughly 20,000 installed centrifuges across its declared sites (Natanz, Fordow, and Isfahan) and an enriched uranium stockpile of approximately 440 kg at 60 percent purity, 184 kg at 20 percent purity, and 6,024 kg at 5 percent purity or below. Following the June 2025 strikes, Iran suspended its cooperation with the IAEA on safeguards inspections required by the Comprehensive Safeguards Agreement.

It remains uncertain whether Iran dispersed key materials and equipment before the strikes. Iranian Foreign Minister Abbas Araghchi claimed that the enriched stockpile was “under the rubble.” In March 2026, IAEA Director General Grossi suggested the majority of the 60 percent highly enriched uranium remained beneath Isfahan and Natanz, though he also said it was “logical to presume” that Iran’s protective measures before the strikes included moving some material. Analysts in recent months have pointed to satellite imagery of convoys transferring blue barrels from Isfahan in June 2025 as evidence of enriched uranium movement. Satellite imagery alone cannot determine what the containers held or where the material went, and drawing firm conclusions risks overstating what open-source intelligence can confirm.

Similar questions arise about the condition and retrievability of material at underground sites. In 2025, the United States used only Tomahawks against Isfahan tunnel entrances, where Iran likely stored over 200 kg of 60 percent uranium, because the GBU-57 bunker buster bomb allegedly could not penetrate Isfahan’s lower levels, which are deeper than Fordow’s. In Operation Epic Fury, the United States targeted underground tunnel entrances to Natanz days after satellite imagery identified vehicle movement suggesting Iran may have accessed the site to recover materials. Even if Iran or another group could reach underground sites through narrow access points, retrieving the gas canisters quickly and safely would be challenging, given strike damage and chemical hazards.

Without a verified accounting of Iran’s enriched uranium, any stockpile disposition framework risks capturing only part of the problem, diminishing the non-proliferation value of any deal regardless of whether the material is down-blended, transferred, or destroyed. Uncovering the status of declared materials would require Iran to first disclose its current locations and then permit third-country excavation and IAEA inventory inspections. Uncertainty about whether Iran dispersed material before the strikes, the strikes damaged it, or it remains retrievable will complicate each step, both technically and politically. Some answers also may not be verifiable at all. If Tehran contends that diverted material is buried and inaccessible, or destroyed entirely, the United States or the IAEA may struggle to verify such claims due to monitoring gaps.

Resolving questions concerning both Iran’s stockpile and its undeclared enrichment capacity with confidence and maintaining a future baseline will ultimately require on-site access to both declared and undeclared locations that only a negotiated inspection regime can provide.

Verification Red Lines for a New Nuclear Deal

By setting back Iran’s nuclear fuel cycle, Operations Midnight Hammer and Epic Fury have provided negotiators with the time required to negotiate a verification regime that closes persistent baseline gaps and covers the breadth of its nuclear program. As nuclear negotiations advance, the U.S. delegation should prioritize three provisions that begin to reconstruct continuity of knowledge after monitoring gaps and preserve it going forward.

First, any nuclear agreement must commit Iran to disclosing its centrifuge production and movement since IAEA monitoring lapsed in February 2021. At a minimum, this should include the location and status of surviving centrifuge stockpiles, production and assembly activities during the monitoring gap, and the construction of undeclared enrichment facilities. Even an imperfect declaration would establish an initial inventory against which future inspections can identify discrepancies and raise the cost of concealment. The framework should then authorize the IAEA to verify the declaration through inspections of centrifuge production and assembly facilities at Karaj, Isfahan, and Natanz, procurement records for components and raw materials, and interviews with relevant personnel.

Second, the deal should require Tehran to declare all nuclear material movement and production since it restricted inspections in June 2025, specifying the location, quantity, and chemical form at each enrichment level. Critically, material at 5 percent and 20 percent warrants close attention alongside the 60 percent stockpile. The JCPOA capped enrichment at 3.67 percent because any material above natural uranium has proliferation risks, and Iran’s accumulation of 60 percent material has not changed that underlying logic. The agreement should also authorize the IAEA to verify Iran’s declaration by granting access to suspected storage and transfer sites and to transport and movement records.

Third, ensuring durable and comprehensive access will require legal instruments beyond the Additional Protocol, whether applied provisionally or ratified. Any framework should enumerate specific sites for inspection from the outset, including Pickaxe Mountain and sites associated with past weaponization-related work (e.g., Taleghan-2, Mojdeh/Lavisan 2, and the Min-Zadayi Complex), to avoid debates over reasonable grounds for access. It must also require Tehran to resume compliance with Modified Code 3.1, a long-standing comprehensive safeguards agreement obligation to submit design information for new facilities once it decides to build one. Substantial nuclear-related sanctions concessions should then be anchored to implementation and compliance with these mechanisms. Initial financial relief could be tied to discrete verification milestones, including access to struck sites, centrifuge declarations, and physical verification of enriched uranium.

Deferring the reconstruction of this baseline accounting of Iran’s nuclear program risks a deal that captures Iran’s formal commitment to a peaceful nuclear program but lacks a monitoring regime to measure compliance. The cost would be persistent ambiguity about Iran’s nuclear program that would repeatedly force Washington back to the same bad choices it faces today: accept Iranian assurances that the United States and IAEA cannot verify or renew military strikes indefinitely. Only an agreement that can reconstruct a working baseline and sustain the access needed to maintain it can break this pattern.