As U.S. and Iranian negotiators return to Islamabad for a second round of talks, it is clear that Washington and Tehran remain far apart on key nuclear issues. With pragmatic diplomacy and the political will to compromise, the United States and Iran can negotiate mutually acceptable provisions to address existing flashpoints, such as the disposition of Iran’s stockpile of uranium enriched to near weapons-grade level and the length of a moratorium on nuclear activities, particularly uranium enrichment. But the nonproliferation value of any nuclear limitations will be minimal without effective verification. Defining the scope of inspections and monitoring mechanisms is arguably the most important component of an effective deal, but it did not appear to be a priority issue during the first round of talks.
As Trump’s team returns to Islamabad, there are certain monitoring provisions that should be non-negotiable priorities for the United States. Other, more detailed provisions may not be feasible to negotiate in a broad framework document under time constraints. If this is the case, the United States should focus on ensuring that, as part of a framework agreement, the International Atomic Energy Agency (IAEA) receives the access and information it needs to determine if any enriched uranium or other key materials were diverted to a covert site and to verify any agreed-upon limits. Unhindered access and timelines for Iran to respond to IAEA inquiries, combined with a multi-year moratorium on proliferation-sensitive activities, would significantly reduce Iran’s proliferation risk. Desirable, but more technically and politically challenging verification provisions, such as verifying the absence of certain weapons-relevant activities, may need to be addressed in subsequent talks. Revisiting monitoring and verification provisions in the future may also be necessary to respond to unanticipated risks and challenges that emerge once on-the-ground assessments of Iran’s nuclear infrastructure begin.
Non-Negotiable: Comprehensive Safeguards with Clarifications
Iran’s resumption of its Nuclear Nonproliferation Treaty (NPT)-required comprehensive safeguards agreement (CSA) should be both non-negotiable and a key priority for U.S. negotiators.
Iran suspended cooperation with the IAEA after U.S. and Israeli strikes on the country’s nuclear facilities in June 2025. Although inspectors were later allowed to access sites unaffected by the conflict in 2025, such as the Bushehr reactor, Iran refused to allow inspectors to visit bombed facilities, citing security concerns, and did not comply with agency requests for updated reporting on the status of Iran’s nuclear materials. As a result, the IAEA assessed in a Feb. 27 report that it lost its continuity of knowledge regarding Iran’s enriched uranium. That report noted that the agency’s “lack of access” to verify the status of the enriched uranium is a “matter of proliferation concern.”
Getting the IAEA back into Iran to begin assessing the status of key nuclear materials, particularly the stockpile of uranium enriched to near weapons-grade, will provide a clearer picture of Iran’s remaining nuclear capabilities and its proliferation risks. Without IAEA inspections, it is challenging to know with any certainty how much of Iran’s nuclear infrastructure remains, how much enriched uranium was destroyed in the strikes, and where the remaining stockpiles of enriched uranium are located.
Iran, for its part, should recognize that failure to meet its legal safeguards obligations will only fuel speculation about illicit activities and increase the risk of a return to conflict. Trump, for instance, used the uncertainties surrounding Iran’s nuclear program as part of his justification for illegal strikes in February.
But simply reiterating that Iran must meet its CSA obligations in a framework deal is insufficient. To ensure full and timely implementation, the United States should push for additional clarifications regarding how Iran implements the CSA.
First, in consultation with the IAEA, the United States and Iran should determine timelines for Iran’s submission of reports on the status and whereabouts of the enriched uranium. Relatedly, negotiators should set a date—ideally as soon as safely possible—to resume IAEA inspections at all nuclear facilities. Agreed-upon timelines will reduce the risk of Iran stonewalling the agency on the provision of information and access, which it has done in the past. If Iran continues to implement an internal process for reviewing IAEA access requests, any agreement should make clear that that process cannot hinder or delay inspections or the provision of reports. The provision of certain forms of sanctions relief or unfreezing of Iranian assets could be contingent upon Iran meeting these IAEA milestones.
Second, any agreement should make clear that the CSA includes the full implementation of Modified Code 3.1—in perpetuity. Modified Code 3.1 amends the CSA’s required notification period for new nuclear facilities. Under the original formulation, a state is only required to submit design information for a new facility to the IAEA six months before the introduction of nuclear materials. Under Modified Code 3.1, a state must provide information about a facility to the IAEA as soon as a decision is made to begin construction. The earlier provision of information allows the IAEA to design a more effective safeguards approach.
Iran claims to have suspended Modified Code 3.1 in February 2021. The IAEA rebuts Iran’s right to do so, arguing that subsidiary agreements cannot be unilaterally suspended. Making clear that a deal will require Iran’s full implementation of Modified Code 3.1 as part of the CSA, and that Tehran cannot renege on this obligation in the future, will help ensure that the IAEA has information about any new nuclear facilities at the earliest possible date. It also reduces the risk that Iran will attempt to reconstruct nuclear facilities in secret and, if caught, try to claim that disclosure to the IAEA was not yet necessary. Relatedly, the agreement should provide clarification about a timeframe for providing design information about nuclear facilities that Iran announced after suspending implementation of Modified Code 3.1 in February 2021.
Given that the IAEA may face unique challenges, including safety issues, arising from the Israeli and U.S. bombings, when it resumes inspections, the United States should agree to fund all additional safeguards expenses and any necessary equipment that the agency might need.
Necessary Supplements to the CSA
Full implementation of the CSA is a necessary step, but it is insufficient to provide assurance that Iran’s nuclear program remains peaceful and mitigate speculation about diversion. The CSA does not give the IAEA the access it needs to determine the status of key components that impact Iran’s proliferation risk, such as the status of centrifuges that were not installed at enrichment facilities. Questions about the status of these non-nuclear components, particularly those relevant to the uranium enrichment program, will exacerbate the risk posed by any gaps in material accountancy. If the IAEA, for example, cannot fully account for all 60 percent enriched uranium and advanced centrifuges, the proliferation threat would be more urgent, as Iran may have diverted both the nuclear material and means to enrich it to weapons-grade levels. To begin to address the gaps in the agency’s understanding of Iran’s full range of capabilities, the IAEA will need access to facilities and information that support Iran’s nuclear program beyond what the CSA facilitates.
Given the importance of assessing Iran’s entire nuclear program, implementation of the additional protocol to Iran’s safeguards agreement should be a top U.S. priority in a deal with Iran. The additional protocol is not legally required by the NPT, but it is an accepted best practice for safeguards. The additional protocol gives the agency access to locations that support a country’s nuclear program but do not house nuclear materials, such as uranium ore concentrate and centrifuge production facilities. It also requires a country to provide the IAEA with additional information about its nuclear program and gives the agency more verification tools. More than 140 states have adopted the additional protocol—requiring it in an agreement does not single Tehran out. From 2016 to 2021, Iran voluntarily implemented the additional protocol as part of the 2015 nuclear deal (but did not ratify it).
Ratification of the additional protocol would be ideal, but it is unlikely that Iran would be willing to take a step now that would be perceived as an irreversible action. The U.S. decision not to require Saudi Arabia to adopt an additional protocol as a condition of nuclear cooperation in the negotiation of a bilateral U.S.-Saudi civil nuclear energy deal could further drive Iranian resistance to ratification of the additional protocol.
Provisional implementation would still provide the IAEA with the access it needs, but again, it could be useful to negotiate specific timeframes for Iran to provide updated declarations about the status of facilities and materials covered by the additional protocol and respond to the IAEA’s questions. Given it has been more than five years since the additional protocol was last implemented and the challenges posed by the destruction of key facilities, reconstructing a baseline of Iran’s nuclear program will likely be a challenging and time-consuming endeavor. Based on consultations with the IAEA, the United States and Iran could agree on an order of prioritization for providing the agency with the necessary information and access. An early emphasis on centrifuge accountancy, for example, could be useful in determining if or how much centrifuge capacity remains in Iran and if/how to monitor the status of those machines in the event of an enrichment pause. Identifying gaps or uncertainties about key inventories early would also be useful in determining what, if any, additional monitoring mechanisms or information may be necessary to further reduce Iran’s proliferation risk.
Avoid Sidelining the IAEA
The IAEA is best placed to monitor and verify the nuclear components of any deal that the United States and Iran negotiate. The agency has the mandate, the expertise, and the history of operating in Iran that will be critical to efforts to begin rebuilding an accurate history of the program that addresses monitoring gaps.
Trump’s negotiators cannot afford to neglect verification requirements in an attempt to hasten progress and/or push for specific safeguards measures that sideline the IAEA in favor of new U.S.-Iran bilateral arrangements. Pursuing separate U.S.-Iran safeguards arrangements also risks undermining the central role of the IAEA and sets a troubling precedent where states may seek to negotiate unique arrangements, rather than adopting widely accepted best practices. Furthermore, it could set up a situation where the IAEA’s efforts are complicated by bilateral mechanisms, making it more challenging to bring Iran into compliance with its NPT-mandated safeguards obligations.
There may be instances where the Trump administration seeks direct involvement in the implementation of certain nuclear issues. Trump, for instance, has suggested that the United States and Iran work together to remove Iran's 440 kg of HEU and transfer it to the United States. Although it is unlikely that Iran would agree to that, if it is part of a deal, the IAEA should be involved in designing and implementing the procedure to ensure accountability.
A more feasible and safer plan would be to down-blend the HEU in Iran to less than five percent enriched uranium. The low enriched uranium could then be stored in Iran (it would pose significantly less risk than the HEU, particularly if stored in powder form and combined with an enrichment moratorium), sold to a third-party country, or possibly owned by Iran but stored at the international fuel bank in Kazakhstan (with the option of return to Iran for fuel fabrication if/when it was necessary for future reactors).
In that scenario, the United States may push for inclusion in the down-blending process or specific monitoring measures (perhaps similar to what the United States and Russia used to monitor down-blending of weapons-grade uranium in the past). The IAEA should still play a key role in designing and overseeing the process to ensure the materials are properly accounted for, and the down-blending can be verified. And any specific U.S. monitoring and verification should be secondary to the IAEA’s procedures.
Empowering and centering the IAEA’s role in monitoring will also help build support for implementing a deal in the United States. Any nuclear agreement with Iran will be subject to the Iran Nuclear Agreement Review Act (INARA), which requires certifications to Congress about the IAEA’s ability to monitor an agreement. Including activities, such as HEU removal and/or down blending, without clearly articulating the IAEA’s role in monitoring and verifying such a process could undermine Congressional support for a deal.
Delegate the Details to Follow Up Negotiations
Trump has made clear that he is not interested in a detailed, technical negotiation and is seeking a framework deal that quickly ends the war. Monitoring and verification, however, should require detailed understandings and timelines in order to prevent Iran from exploiting loopholes or drawing out the safeguards process, as it has in the past.
One option could be for any U.S.-Iran deal to include broad principles outlining monitoring and verification, such as stating that Iran will fully implement the CSA and additional protocol, with the timeline for certain activities to be determined in consultations with the IAEA. Then, the IAEA, Iran, and the United States could negotiate the more specific procedures and timeframes in line with the broad parameters outlined in the framework agreement. This could provide greater assurance that a deal will be verifiable and that the agency will have the information and access it needs. To incentivize and ensure follow-up talks on the specifics—and perhaps later more intrusive provisions—benefits could be tied to reaching and implementing a more detailed monitoring arrangement.
Monitoring and Verification Flexibility
Given the uncertainties about the status of Iran’s nuclear materials and key technologies, the monitoring regime should be designed to be flexible and responsive to issues that may arise. For example, accountancy gaps in the stockpile of 60 percent enriched uranium could necessitate more intrusive surveillance of remaining centrifuges or shorter timelines for Iran to accept IAEA requests for access to sites not covered by the additional protocol or CSA.
Including language for follow-on talks about monitoring and verification could also create space for defining more intrusive monitoring mechanisms that are desirable but may be too complicated to address in a shorter framework agreement to end the war. It would be useful, for instance, to consider how to monitor the absence of certain weaponization-related activities or, alternatively, develop a process for investigating accusations of such work. It is unlikely, however, that a framework agreement focused on ending the war will go into that level of detail or that negotiators will have the time and expertise to think through those questions.
Prioritizing Verification
From a nonproliferation perspective, negotiating an effective verification regime will be crucial for the success of any nuclear agreement. Not only is an intrusive monitoring regime necessary to ensure that Iran adheres to the nuclear limitations of a deal, but it also provides greater assurance that any past (or future) diversion of materials and technologies will be detected. Effective verification also benefits Iran, including by reducing the risk of conflict re-erupting based on mischaracterizations of the nuclear program. It must be a key priority in negotiations going forward.—KELSEY DAVENPORT, director for nonproliferation policy