Will everyone want a bomb?
The nuclear Non-Proliferation Treaty is reviewed every five years by its 191 member states, with the aim of producing a consensus document on compliance and next steps across the treaty's three pillars: disarmament, non-proliferation, and peaceful nuclear energy. The last three conferences (2005, 2015, 2022) failed to produce one. The 2026 Review Conference opens in New York on April 27, and the five nuclear-weapon states that anchor the treaty are, in various ways, actively weakening it. The question for the four weeks of negotiation ahead is whether the basic exchange at the heart of the NPT, restraint in return for security and disarmament, still holds.
The US and Israel have struck Iranian nuclear facilities twice in nine months. Russia has revised its nuclear doctrine to lower the threshold for use. China is covertly expanding its warhead production infrastructure. The last bilateral US-Russia arms control treaty expired in February with no successor. And the three-year preparatory process for this conference failed to agree on a single recommendation. Vietnam's Ambassador Do Hung Viet chairs the conference. He inherits a process in which all five permanent Security Council members sent sub-ambassadorial delegations to the final preparatory session.
The P5
The US under the Trump administration set a two-month deadline for Iran nuclear negotiations in March 2025, then ran five rounds of indirect talks in which its position shifted between demanding full dismantlement of Iran's enrichment programme and signalling flexibility on limited enrichment, imposing new sanctions between rounds. When the deadline expired, Israel struck Iranian nuclear facilities and military leaders. The US struck separately, hitting Fordow, Natanz, and Isfahan. A second round of talks opened in February 2026, and within 48 hours of what Oman called "significant progress," the US and Israel launched a second wave, killing Khamenei and senior commanders. Washington has also signalled it will oppose any reference to UN sustainable development goals in the conference document, a position that could collapse negotiations on its own.
Russia blocked the 2022 conference over language about its occupation of the Zaporizhzhia nuclear power plant. Its nuclear posture has hardened since the full-scale invasion of Ukraine, with revised doctrine and periodic signalling directed at NATO. Whether Russia engages constructively in 2026 is an open question.
China was disruptive at the 2025 preparatory session, blocking consensus on transparency and reporting while demanding that states under US nuclear umbrellas report on extended deterrence arrangements. China is expanding its arsenal, with satellite imagery confirming new warhead production facilities in Sichuan province.
Europe is largely marginalised. The UK and France maintain relatively small arsenals and have generally supported the treaty architecture, but neither has shaped a substantive initiative for this conference. France triggered the European snapback mechanism on Iran sanctions, yet was excluded entirely from the US-Iran negotiations. Europe's primary concern (Russian nuclear threats) does not align with Global South priorities around disarmament and the precedent set by the Iran strikes. The EU retains diplomatic weight but has no clear position.
The bargain
The NPT rests on an exchange: non-nuclear-weapon states forgo the bomb, and in return receive security assurances, access to peaceful nuclear technology, and progress toward disarmament by those that already have weapons. The recent record available to any government assessing that exchange is not reassuring. Ukraine gave up its nuclear weapons in return for security assurances and was invaded. Libya dismantled its programme and was overthrown with NATO support. Iran constrained its enrichment under the JCPOA; the US withdrew and later struck its facilities. North Korea kept its weapons and has not been attacked.
Each case is different but this is a pattern. George Perkovich of Carnegie has noted that the cost-benefit signal is now clear enough for any government to read. States allied to the US wonder whether Washington will fight for them. States that are adversaries have seen what relying on the NPT bargain gets you.
No new state has acquired nuclear weapons as a result. But the political debate is shifting. South Korea's Foreign Minister published an essay on April 1 arguing that his country's commitment to the NPT is a strategic choice, not a legal constraint, and that nuclear armament would undermine rather than strengthen South Korean security. That a sitting Foreign Minister felt compelled to make this case publicly tells you something about the pressure building in Seoul, where polling has consistently shown majority support for an independent nuclear capability. Similar, if less advanced, discussions are emerging in Nordic countries. These are US allies questioning extended deterrence, not adversaries.
India, Pakistan and Israel sit outside the treaty entirely and are barely discussed in the review process, despite collectively holding several hundred warheads. Israel's strikes on Iran were conducted by a state that has never joined the NPT or declared its arsenal. The Arab states have long tied support for conference outcomes to progress on a Middle East WMD-free zone. After the strikes, that position will harden.
The erosion is not confined to nuclear governance. International humanitarian law, the Geneva conventions, and institutions of international criminal justice are under comparable pressure from states that built these frameworks and now treat compliance as selective. The NPT is one part of a wider pattern in which legal constraints are loosening faster than the institutions responsible for them can respond.
What to watch
Do Hung Viet's task is merely to determine whether any outcome is possible.
Iran will be the first test. A majority of states will push for language condemning the US and Israeli strikes, which the US will block. This could replicate the dynamic that sank the 2022 conference. If Iran moves toward withdrawal, it would overshadow everything else.
A deeper tension runs through the preparations. One camp argues the treaty is sound and the problem is non-compliance, particularly by the nuclear-weapon states, with the answer being stronger political pressure and parallel instruments like the Treaty on the Prohibition of Nuclear Weapons. Another argues the review process itself needs adaptation, including alternative outcome formats and voting procedures already available under the rules but never used. The nuclear-weapon states have little interest in either route. A process that constrains others, absorbs diplomatic energy, and produces no binding accountability for their own commitments is one they can live with. A failed consensus suits them better than a recorded vote.
A fourth consecutive failure would not be surprising. But it would confirm that a process absorbing years of preparation and weeks of negotiation cannot produce outcomes even when nuclear risk is at its most acute in decades. If the review cycle cannot deliver, governments will look elsewhere: to bilateral deals, minilateral arrangements, or their own deterrence capabilities. The longer it fails, the more those alternatives stop looking temporary and start becoming the architecture.
