Abstract
For nearly eight decades, the absence of nuclear war has depended not solely on the effectiveness of nuclear deterrence but also on institutional safeguards, political restraint, and, at times, chance. As arms control arrangements erode, geopolitical competition intensifies, and emerging technologies introduce new risks, reliance on deterrence alone appears increasingly fragile. Nevertheless, universal nuclear disarmament remains achievable not through immediate multilateral abolition, which remains politically unrealistic in the near term, but through a carefully sequenced process of verified and incremental reductions grounded in strategic realism. A five-stage framework encompassing renewed US–Russia arms control, China’s integration into strategic stability arrangements, governance of artificial intelligence in nuclear decision-making, regional risk reduction measures, phased global reductions, and institutional reform could be implemented. By linking strategic stability, emerging technologies, and multilateral governance, a practical pathway toward universal nuclear disarmament in an increasingly multipolar nuclear order could be feasible.
Introduction
The case for nuclear disarmament has never rested on sentiment alone. It rests on arithmetic. The world currently possesses approximately 12,187 nuclear warheads. A fraction of these, if used, would trigger atmospheric, agricultural, and civilisational consequences that no emergency response architecture could meaningfully address. The question confronting the international community is therefore not whether nuclear disarmament is desirable—it is whether it is possible to chart a politically realistic path towards achieving it given the constraints of the present strategic environment.
The 11th Review Conference of the Treaty on the Non-Proliferation of Nuclear Weapons (NPT), convened to review implementation of the Treaty across its three mutually reinforcing pillars—nuclear disarmament, nuclear non-proliferation, and the peaceful uses of nuclear energy—concluded without a consensus outcome document, marking the third consecutive Review Conference to do so. The conference’s outcome reflected a broader lack of political will among nuclear-armed states to implement their legally binding disarmament commitments in good faith. Conference President, Ambassador Do Hung Viet of Vietnam, captured the moment with characteristic precision: “If we cannot agree on which road to take, we will never arrive at our destination.” In this regard, a politically realistic pathway to universal nuclear disarmament lies in a sequenced approach to verified reductions. A five-stage framework that begins with restoring strategic stability among the major nuclear powers, extends to governance of emerging technologies and regional risk reduction, and culminates in phased global reductions supported by strengthened multilateral institutions could be developed.
Step One: Restore the Bilateral US–Russia Arms Control Foundation
Russia possesses approximately 5,420 warheads and the United States approximately 5,042—together accounting for nearly 90 percent of global nuclear capacity. With the expiration of the New START Treaty in February 2026, no legally binding limits now constrain either country’s strategic nuclear forces. Prior to the treaty’s expiry, Russia proposed that both sides continue voluntarily observing its central numerical ceilings of 1,550 deployed strategic nuclear warheads, 700 deployed strategic delivery systems, and 800 deployed and non-deployed launchers till February 2027 while negotiating a successor agreement. The US did not formally respond to the proposal, instead signaling its preference for negotiating a broader arms control framework that would extend beyond the existing bilateral architecture and include China. This is the most treacherous structural gap in the contemporary nuclear order.
While political differences remain substantial on both sides, sustained risk-reduction measures and strategic dialogue could gradually create the conditions for future arms control negotiations. The first concrete step is not a new treaty but strengthening military-to-military communication channels, maintaining and enhancing existing nuclear crisis hotlines, and resuming regular strategic dialogue. Their purpose is not disarmament but the prevention of accidental escalation, creating space within which formal negotiations can resume. The second step is an interim restraint arrangement: a political commitment by both sides to voluntarily maintain the former New START numerical ceilings — 1,550 deployed strategic warheads per side while successor negotiations proceed. The third step is a successor agreement that transcends Cold War categories, addressing not only deployed strategic warheads but also non-strategic, shorter-range tactical nuclear weapons, hypersonic delivery systems, missile defence architectures, and AI-assisted targeting technologies. Phased reductions — beginning with a deployment freeze, then reserve warhead reductions, then tactical systems that are more politically durable than deep immediate cuts and create the conditions for deeper structural change over time.
Step Two: Integrate China into a Trilateral Stability Framework
China has expanded its arsenal to approximately 600–620 warheads, is projected to exceed 1,000 operational warheads by 2030, has constructed new intercontinental ballistic missile (ICBM) silo fields, completed a full nuclear triad, and is deploying hypersonic glide vehicles. While the US and Russia continue to possess the largest arsenals, China’s rapid quantitative and qualitative expansion is reshaping strategic deterrence dynamics. Bilateral agreements that exclude China risk leaving one of the three principal nuclear powers outside future arms control constraints, thereby reducing the long-term effectiveness and sustainability of any strategic stability framework.
China has consistently declined to participate in trilateral nuclear arms reduction negotiations with the US and Russia, arguing that doing so would institutionalise permanent strategic inferiority given the significant disparity in nuclear arsenals. The realistic response is therefore not to demand immediate participation on terms Beijing is unlikely to accept but to design a sequenced pathway for engagement. The first stage could focus on limited transparency without immediate numerical caps. Rather than full declaratory disclosure, China could provide broad estimates of its warhead stockpile and delivery systems as part of reciprocal confidence-building measures involving the US and, where appropriate, other nuclear-armed states. Such transparency would need to be accompanied by assurances that it would not be used to justify further quantitative or qualitative arms buildups, thereby addressing Beijing’s long-standing concerns over strategic vulnerability. Although the contemporary nuclear order is increasingly shaped by overlapping deterrence relationships, including the US–China, China–India, and India–Pakistan dyads, the expanding strategic interaction among the US, Russia, and China makes trilateral engagement an increasingly important component of any future global arms control framework.
The second stage involves formal strategic stability dialogues covering missile defence, hypersonic systems, and AI in nuclear command structures. The third stage envisages a trilateral freeze agreement preventing unconstrained expansion by the US, Russia, and China. Such an arrangement would need to take account of the security concerns of India, which may be reluctant to support a framework perceived as entrenching China’s numerical advantage. Consequently, any freeze should be accompanied by broader strategic stability dialogues and confidence-building measures that reassure regional actors and preserve the possibility of future multilateral participation. The fourth converts the freeze into proportional reduction commitments applicable to all three powers simultaneously. China’s formal No First Use (NFU) provides a useful starting point for discussions on reducing nuclear risks. However, broader doctrinal convergence will remain difficult unless the US and Russia are also willing to engage, particularly given the current political positions of their leadership.
Step Three: Establish Universal Norms on AI in Nuclear Command Systems
The disarmament agenda cannot be limited to warhead numbers. The progressive integration of AI into nuclear command, control, and communications (N3C) systems introduces a qualitatively new and underregulated category of risk. AI-enabled early warning systems compress human decision-making timelines. Machine-assisted targeting technologies reduce the space for political judgment during crises. The danger is not that states will formally delegate launch authority to machines — it is that excessive dependence on AI-generated assessments during acute geopolitical tension will produce escalation before political leaders have fully understood the situation. A manipulated dataset, sensor malfunction, or algorithmic error could trigger a sequence of events that human judgment alone would have arrested.
India articulated these concerns at the High level Segment of the Conference on Disarmament (CD) in February 2026, emphasising the growing risks posed by AI, cyber capabilities, and autonomous systems in the strategic domain. The 11th NPT Review Conference’s draft outcome document had included language on maintaining human control over nuclear weapons—language dropped under pressure from nuclear-armed states. The concrete step required is the expansion of United Nations General Assembly (UNGA) Resolution 80/23 to explicitly govern AI within nuclear decision-making. This effort should complement existing international initiatives, including the Responsible AI in the Military Domain (REAIM) process, by developing specific norms for the use of AI in nuclear command, control, and decision-making. The human-in-the-loop standard must be codified as a universal norm within all nuclear doctrines. States must simultaneously commit to prohibiting cyberattacks targeting nuclear command infrastructure, including malware insertion, data poisoning, and algorithmic manipulation.
Step Four: Modernise Regional Nuclear Agreements
Regional nuclear flashpoints matter because they shape the wider strategic environment in which the US, Russia, and China negotiate. Instability in South Asia, the Middle East, and East Asia and Europe can reinforce proliferation pressures, reduce confidence in disarmament, and make global arms control more difficult to sustain. For that reason, any credible framework for great-power dialogue must be accompanied by regional risk-reduction measures. In addition, great-power frameworks will lack credibility if regional nuclear environments are left unaddressed.
In South Asia, India and Pakistan maintain one of the world’s most operationally precarious nuclear relationships—characterised by compressed decision timelines, recurring military crises arising from cross-border terrorism, and the growing deployment of unmanned aerial systems that blur the boundary between conventional provocation and strategic threat. The existing India-Pakistan Non-Nuclear Aggression Agreement must be modernised to include prohibitions on cyberattacks against nuclear infrastructure, restrictions on hostile drone operations near strategic installations, and dedicated crisis communication channels for drone and cyber incidents.
In the Middle East, Israel’s policy of deliberate nuclear ambiguity cannot be permanently exempt from a universal disarmament architecture. The realistic entry point is a graduated transparency process: formal acknowledgment of nuclear capacity, followed by limited engagement with international monitoring frameworks, followed eventually by participation in regional strategic stability dialogues.
In East Asia, North Korea’s continued expansion—solid-fuel ICBMs, submarine-launched ballistic missiles, and domestic legislation permitting preemptive nuclear use—combined with intensifying proliferation debates in South Korea and Japan, signals a deteriorating regional environment. The credible interim pathway involves phased transparency measures from Pyongyang, reciprocal security guarantees, and graduated sanctions relief, with China and Russia playing a more constructive facilitative role than they have to date.
In Europe, France assumes a more prominent role in European nuclear deterrence amid growing uncertainty over the long-term credibility of US security guarantees. Therefore, regional stability should be strengthened through renewed dialogue on intermediate-range missiles, non-strategic nuclear weapons, missile defence, and transparency measures.
These efforts should complement, rather than substitute for, broader US–Russia strategic stability and future trilateral arms control initiatives.
Step Five: Implement a Phased Global Reduction Architecture
Once bilateral, trilateral, AI governance, and regional foundations are in place, the framework moves toward formal quantitative reductions across all nuclear-weapon states. The initial benchmark should not be abolition but minimum deterrence — each state reducing its arsenal to approximately 50 to 60 warheads under strict verification regimes, which should include regular declarations, on-site inspections, satellite monitoring, and an independent international verification mechanism drawing on IAEA expertise and established arms control verification practices. This threshold preserves functional deterrence – which retains a credible second-strike capability sufficient to deter nuclear attack without maintaining excessive nuclear arsenals. As stockpiles approach lower levels, universal NFU adoption becomes structurally more favorable. China and India already maintain formal NFU postures, demonstrating their operational compatibility with functional deterrence. Extended nuclear deterrence arrangements—particularly within the North Atlantic Treaty Organization (NATO)—cannot be dismantled abruptly. Any reductions must be accompanied by stronger conventional deterrence to avoid creating uncertainty that could incentivise further proliferation. This process will also require greater political cohesion within NATO, which is currently facing significant internal strategic and political challenges.
Conclusion
For nearly eight decades, nuclear war has been avoided not through the perfection of deterrence but through institutional safeguards, political restraint, and chance. The framework proposed here does not require states to abandon deterrence overnight only to take the first step. Restore US-Russia arms control. Integrate China into a trilateral stability framework. Strengthen norms on AI in nuclear decision-making and modernise regional agreements in South Asia, the Middle East, and East Asia. Move toward verified, phased reductions to minimum deterrence thresholds.
Each step demands sustained political will that has often been lacking in recent years. Yet the alternative is not a stable status quo but the continued erosion of existing arms control frameworks and growing strategic uncertainty. The pathway to universal nuclear disarmament is neither simple nor immediate, but it remains achievable through gradual, verifiable, and politically realistic measures. The challenge is whether states are prepared to take the necessary steps to follow it.
Nichole Ballawar is a public policy and strategic affairs analyst who previously served as a consultant in the Ministry of External Affairs and the Ministry of Heavy Industries, Government of India. His research focuses on nuclear policy, arms control, disarmament, emerging technologies, and international security.
All views expressed, as well as any errors, remain the sole responsibility of the author.