Following more than half a century of arms control treaties that provided at least some guardrails on nuclear competition, today’s expiration of New START confirms that this architecture has largely unraveled. Over the coming year, nuclear weapons policy is likely to be shaped less by dramatic doctrinal shifts than by the cumulative effects of institutional strain, new proliferation risks, and the challenges posed by integrating emerging technologies into the nuclear domain.
Rather than a single inflection point, policymakers are likely to confront a series of pressures that together are reshaping the strategic environment. The erosion of arms control frameworks, the widening of proliferation risk beyond the traditional countries of concern, and the growing entanglement of nuclear forces with cyber, space, and artificial intelligence systems each pose distinct challenges.
Taken together, they point to a nuclear order that is becoming more fragmented, less predictable, and increasingly difficult to govern through existing institutions.
The End of Arms Control, For Now
New START’s expiration this week marks the first time since the early 1970s that the world’s two largest nuclear arsenals are unconstrained by a formal, verifiable agreement. With nuclear relations between the United States and Russia at their lowest ebb in decades, and Moscow having suspended treaty inspections and data exchanges well before expiration, there is little prospect of either an extension or a successor framework in the near term (despite recent efforts by Moscow to rekindle discussions). Efforts to bring China into a trilateral arms control arrangement—long a U.S. objective—are likewise stalled, as Beijing continues to reject participation in formal numerical limits while quantitatively increasing and qualitatively broadening its nuclear forces.
In broader terms, arms control is unlikely to disappear entirely, but it will take on a thinner, more fragmented form. Limited confidence-building measures—such as unilateral limits on arms numbers, crisis communication channels, notifications of major exercises, or declaratory dialogue on nuclear doctrine—may still be achievable through bilateral contacts or multilateral forums, particularly among the five official nuclear weapon states (P5). Yet these mechanisms lack the binding force, verification, and predictability that defined earlier arms control regimes—with no guarantee that the other side will pick up the proverbial red phone.
At the same time, all nuclear-armed states are pressing ahead with modernization programs that embed new delivery systems, updated command-and-control architectures, and increasingly sophisticated dual-use technologies into deterrence postures originally designed for a bipolar Cold War context. China’s buildup of silo-based ICBMs, expansion of its ballistic missile submarine force, and diversification of delivery options underscore how deterrence dynamics are shifting toward a more complex, multi-actor nuclear environment that might make future arrangements difficult to achieve.
The cumulative effect is not necessarily a simple return to Cold War–style arms racing, but a more opaque nuclear order in which signaling becomes harder to interpret and risk management rests increasingly on national judgment rather than shared rules. The absence of sustained dialogue among the three major nuclear powers magnifies the danger that modernization, force expansion, or routine military activities will be misread as escalatory intent. In such a setting, stability depends less on formal ceilings and more on restraint, communication, and institutional resilience—precisely the areas where trust is thinnest and political incentives for cooperation are weakest.
Nuclear Proliferation
Related to the challenges faced in the arms control domain, the proliferation challenge facing the global nuclear order is increasingly defined not by sudden breakouts (à la North Korea), but by the steady spread of nuclear latency and hedging behavior among nuclear-capable states. Iran remains the most acute near-term concern. Tehran’s expanding enrichment capacity, accumulation of near-weapons-grade material, and reduced international monitoring have compressed the technical timeline for weaponization, evan as Iran’s leaders continue to signal a willingness to engage in diplomacy—amid significant domestic political unrest.
The proliferation challenge is also widening to include advanced industrial states that remain formally committed to nonproliferation but are quietly reassessing the durability of their restraint. Heightened great-power rivalry, the visible erosion of arms control outlined above, and growing doubts about the credibility of U.S. extended deterrence commitments are prompting renewed debate within U.S. alliances. In East Asia, both Japan and South Korea possess the technical expertise, industrial infrastructure, and fissile material access required for rapid nuclearization. While neither government has moved toward weaponization, domestic discussions about nuclear latency, indigenous deterrent options, and alternative basing arrangements have become more politically salient amid North Korea’s advancing capabilities and China’s expanding nuclear forces.
In Europe, Russia’s war in Ukraine and its repeated use of nuclear signaling during the conflict have similarly unsettled long-standing assumptions about the relevance of nuclear capabilities in crisis. States such as Poland remain firmly embedded within NATO, yet have begun to explore long-term security options that reflect heightened threat perceptions and uncertainty about future U.S. commitments to collective security (made more acute by recent events in Davos and U.S. claims to Greenland).
These debates do not signal imminent proliferation, but they do normalize nuclear hedging as a rational response to systemic stress. The danger for the nonproliferation regime lies precisely in this normalization: a world in which more states retain the technical option to proliferate, shorten decision timelines, and treat nuclear restraint as conditional rather than categorical. Over time, these dynamics risk hollowing out the regime from within—not through overt defection, but through the gradual erosion of norms that once made abstention politically unthinkable.
Emerging Capabilities
A third defining feature of the year ahead (and focus of my own Lab’s work) will be the growing salience of cross-domain capabilities—particularly the interaction of nuclear forces with advances in AI, cyber operations, space systems, and precision conventional strike. Rather than altering declaratory nuclear doctrine, these technologies are reshaping how states perceive threats, manage crises, and operate under conditions of uncertainty. Governments are already investing heavily in intelligence, surveillance, and reconnaissance systems designed to improve strategic early warning and target detection, while also modernizing nuclear command, control, and communications (NC3) to ensure resilience against disruption. These efforts reflect a shared concern that future crises will unfold more rapidly, with compressed decision timelines and greater volumes of ambiguous or conflicting information.
At the same time, the integration of digital technologies into nuclear-adjacent systems introduces new risks alongside their operational benefits. AI-enabled decision-support tools promise faster analysis and pattern recognition, but they also raise questions about automation bias, model opacity, and the amplification of worst-case assumptions during crises. Cyber operations targeting dual-use infrastructure—such as early-warning sensors, satellites, or communications networks—also risk blurring the line between conventional and nuclear escalation, particularly if states cannot easily distinguish between espionage, preparation of the battlefield, and disabling attacks. Space systems, long treated as enabling infrastructure, are increasingly viewed as potential points of contestation (not least given reports of Russia’s pursuit of a nuclear weapon in space), further complicating escalation dynamics when interference with satellites could be interpreted as a prelude to broader conflict.
These challenges are compounded by the fact that most existing arms control and risk-reduction frameworks were not designed with cross-domain concerns in mind as they, by definition, tend to narrowly circumscribe the technologies under consideration. Indeed, there are few shared rules governing cyber operations against NC3, the military use of AI in strategic contexts, or the protection of space-based early-warning assets. As a result, states are navigating these issues largely through national doctrine, internal safeguards, and informal exchanges rather than binding agreements. Over the coming year, the most consequential developments in this space are therefore likely to occur quietly: through state-level decisions about human–machine interaction, thresholds for automation, redundancy and fail-safe design, and restraint in targeting practices.
While incremental, these choices will play an outsized role in shaping whether emerging technologies ultimately stabilize deterrence—or make an already fragile nuclear order more prone to miscalculation.
A New World Order
Taken together, these developments push nuclear weapons policy debates away from headline questions about warhead numbers or treaty ceilings and toward quieter—but consequential—discussions about operational safeguards, human-machine interaction, and the governance of enabling technologies. In that sense, the most important nuclear policy developments in the coming year may occur not through new agreements or doctrines, but through incremental decisions about how nuclear systems are embedded within broader military and technological ecosystems in both nuclear and non-nuclear weapons states.
– Andrew W. Reddie, Published courtesy of Just Security.
