Amid the rapid expansion of AI and data centers, and as climate commitments drive demand for clean energy, nuclear power is increasingly seen as a reliable source to fill gaps and complement the intermittent output of solar and hydropower. This renewed interest is also spurred by nations’ pursuit of greater energy self‑reliance, even as energy and nuclear infrastructure become consequential targets in armed conflict—disrupting electricity supplies, undermining military capabilities, and serving counter‑proliferation objectives. Small modular reactors have re‑emerged as a public‑relations vehicle for the nuclear industry, offering a narrative that smaller, more flexible reactors can make nuclear power more accessible and easier to deploy than large conventional plants.
Beyond the U.S. Department of Defense’s push to rapidly develop and field small modular reactors—potentially bypassing traditional Nuclear Regulatory Commission processes—and to leverage its substantial budget to jump‑start the industry commercially, nuclear enthusiasm is evident worldwide. Renewed investments across Europe and Asia—from Seoul’s deepening nuclear cooperation in the East, to Saudi Arabia’s push into civil nuclear in West Asia, to amendments of India’s decades‑old civil nuclear laws in South Asia, and a growing roster of countries seeking 123 agreements with the United States—testify to a resurgence spanning both established powers and aspirant newcomers. This momentum extends beyond conventional reactors, with nearly 80 small and advanced reactor designs under development globally.
Even if still largely “paper tigers,” the scale of investment and policy momentum point to one broader implication: regardless of whether these new‑age reactors eventually become commercially viable, the current small modular reactor push will have ramifications extending far beyond the technology itself—strengthening overall confidence in nuclear power and potentially driving greater deployment of conventional reactors. However, this apparent nuclear renaissance is outpacing the rules designed to safeguard it. As nuclear reactors—traditional and advanced—potentially proliferate, new governance challenges arise. While not entirely new, these concerns are compounded by evolving deployment patterns across civilian, industrial, and military applications and the changing nature of battlefields.
The central dilemma is whether these facilities become sitting ducks—exposed to attack by virtue of their location, mobility, or proximity to military assets—or strategic shields whose potential radiological consequences and wider escalation risks constrain an adversary’s military choices.
The governance challenge, therefore, is to get ahead of the technology rather than retrofitting rules after deployment patterns have become entrenched.
Small Is Not Always Simple
Part of the appeal of small and advanced reactors rests on the premise that these new‑age reactors are marketed with the potential to mitigate conventional nuclear safety risks through smaller reactor cores and passive cooling systems. Their design and deployment options—such as underground placement to enhance protection against attacks, clustered deployment to achieve greater aggregate energy output, and co‑location alongside critical defense infrastructure—also contribute to their purported utility. These features, however, also introduce a distinct risk profile that requires tailored governance arrangements. Underground installations, for instance, may face issues related to cooling disruptions and management of heat decay after automatic shutdowns, while groundwater intrusion through flooding or geological shifts further complicates emergency access and inspection. Clustered deployments, meanwhile, may introduce interdependencies between modules, leading to cascading safety risks, where a failure or disruption affecting one unit could compromise the operation or safety of others.
These challenges are compounded by the economics of deployment. Small modular reactors in early deployment phases are already more expensive and slower to commercialize until manufacturing efficiencies and standardization are achieved. These constraints become critical when assessing initial capital costs against electricity output. This creates a chicken‑and‑egg dilemma where large‑scale deployment is needed to realize economies of scale, yet achieving that scale depends on sufficient demand, financing, and commercial viability. Near‑term applications are therefore expected to be concentrated in more niche settings where reliable, localized power is particularly valuable for remote greenfield and brownfield sites.
This is critical for governance because the more specialized the deployment landscape becomes, the less straightforward it is to rely on existing assumptions about where nuclear facilities will be located, how they will be protected, and how they may feature in crisis scenarios.
From Civilian Asset to Strategic Liability
Many of these small and advanced reactors are being developed with the prospect of siting alongside military installations, given the military’s need for a separate, secure power grid capable of sustaining operations, especially where civilian grids risk being disrupted by adversarial attacks.
The Department of Defense, for example, has long explored the feasibility of mobile microreactors for remote military operations through initiatives such as the prototype Project Pele in 2019 and its follow‑on, Project Janus, which envisages the deployment of commercially owned and operated microreactors (1–20 megawatts) that can be regulated by the U.S. Army at up to nine military bases. Similarly, microreactors are also being proposed for co‑location with data centers and other critical infrastructure, reducing dependence on distant transmission networks. Now, if such a reactor located within a military base or alongside a data center directly supports military functions, its status under the laws of armed conflict becomes more complex, and in some circumstances may satisfy the definition of a military objective.
Several provisions, including Article 56 of Additional Protocol I to the Geneva Conventions, provide nuclear electricity‑generating stations with special protection by prohibiting attacks on such facilities and on military objectives located in or around their vicinity, particularly when an attack may cause the release of dangerous forces and consequent severe losses among the civilian population. Such special protection, however, ceases where the facility provides electric power in regular, significant, and direct support of military operations and an attack is the only feasible way to terminate that support.
There also remain divergences in state practice of such provisions. The United States, which is not a party to Additional Protocol I, accordingly recognizes that attacks on nuclear power plants may be permissible under applicable rules of distinction and proportionality as per the Department of Defense’s Law of War Manual, while Russia, despite being a party, continues attacks on nuclear infrastructure. Rule 42 of the customary international humanitarian law framework of the International Committee of the Red Cross also stipulates that attacks on installations containing dangerous forces, as well as on military objectives situated within or near them, must be conducted with particular caution. In either framework, the potential for radiological consequences remains highly relevant to the legal assessment.
At the operational level, the concern lies in the vulnerabilities of reactors deployed in conflict zones. While these reactors are emphatically marketed to be designed with enhanced safety features, they do not completely eliminate the possibility of a radiological release. The scale of fallout depends on the reactor’s design, fuel inventory, operating conditions, and the nature of the damage, and cannot be juxtaposed against conventional examples. But even a comparatively limited fallout would carry political and legal consequences disproportionate to its physical effects and set a problematic precedent of exposing nuclear facilities to conflict‑related risks.
Although such attacks remain exceptional in the near term, their growing plausibility cannot be dismissed. Moreover, even when not intended targets, such reactors risk becoming collateral and sustaining accidental damage amid a changing battlefield that is increasingly characterized by the growing use of drones and uncrewed systems and compressed decision‑making timelines—particularly when they are located near military or other critical infrastructure. Several documented cases illustrate the deliberate threat drones pose to critical infrastructure. However, reports of incidental drone damage to a power reactor in a conflict zone remain scarce. This may also be attributed to a small sample size, as the number of reactors in active conflict zones is limited.
At the systemic level, the concern shifts from individual vulnerabilities to the cumulative risks of wider deployment. Their wider proliferation—whether militarily or commercially—will lead to a larger number of dispersed nuclear facilities potentially becoming exposed to conflicts. In effect, these reactors could become vulnerable sitting ducks, creating new pathways for crisis instability.
On the flip side, dispersed microreactors could reduce dependence on centralized power stations and transmission networks, limiting the disruption caused by attacks on major grid infrastructure. Yet this resilience may come at a cost.
The risk is neither uniform nor inherent to the technology. It depends on where and why reactors are sited as civilian and military functions overlap, creating vulnerabilities that are difficult to assess as lawful targets.
Portability Without Predictability
The potential transportability of advanced microreactors introduces another governance challenge, conflating physical mobility, information uncertainty, and strategic ambiguity.
To illustrate, earlier this year, the U.S. Departments of Energy and Defense transported a Valar Atomics Ward 250 microreactor by C‑17 from California to Utah without fuel—marking the first air transport of a nuclear microreactor in the country. The exercise demonstrated that compact nuclear power systems designed for expeditionary or forward‑operating environments may eventually enable modular transport and redeployment between sites, even if such movements do not become routine practice. Additionally, the U.S. Navy is set to develop a 1‑megawatt microreactor prototype with Radiant Industries at Naval Weapons Station in Indiana to support critical installation infrastructure. The technology is intended to be portable across land, sea, and air. Although these instances remain peacetime demonstrations of transportability, their potential implications for deployment in conflict settings merit attention—before operational practices and doctrines evolve.
The prospect of transportable nuclear power is not confined to U.S. initiatives. China’s pursuit of such reactors ranges from 10‑megawatt vehicle‑mounted reactor prototypes, dubbed “nuclear power bank,” to longer‑term proposals for nuclear‑powered tanks equipped with a long‑range electromagnetic railgun.
If these systems become operational and are eventually adopted across countries, their mobility could further complicate existing nuclear confidence‑building measures, such as those in South Asia, which rely on the periodic identification and exchange of information on protected nuclear facilities by making the location of certain reactors less predictable. Would a mobile reactor be declared as an installation in its own right, or would its declaration attach to its normal or host facility? Would a change in location require notification outside the annual exchange cycle? And, more fundamentally, should future confidence‑building measures account for the movement of protected nuclear facilities rather than assume that their location remains fixed? Transportability could increase the possibility of miscalculation during conflict, particularly if the location or status of a reactor is not known to an adversary.
These technologies are at varying stages of development, with considerable uncertainty surrounding their eventual deployment and applications. These questions, hence, are not confined to any single reactor design. They consider the broader governance and security implications that may emerge as such technologies develop and potentially enter new operating environments.
The security rationale for mobile microreactors, however, is not without merit. At forward military bases, reliable nuclear power could reduce dependence on diesel generators and fuel convoys, which are themselves vulnerable to attack and have historically exposed personnel to significant risks. This is not an entirely new military ambition either—Cold War precedents include the U.S. Army’s PM‑2A reactor in Greenland, and the Soviet Union’s Pamir‑630D mobile reactor, among others. Although these initiatives were discontinued amid operational and safety challenges, advances in reactor design and passive cooling may have improved prospects. Large‑scale deployment in active conflict zones, nevertheless, remains unproven. Microreactors, therefore, while mitigating existing vulnerabilities, may introduce new ones.
Sitting Ducks to Strategic Shields?
The prospect of mobility also changes how reactors might be used in a conflict. For example, a nuclear power source that is not permanently tied to a fixed installation could, in principle, be deployed within a theater of war, carrying strategic implications beyond its intended power‑generation role. Hypothetically, a commander may eventually be able to airlift a reactor into an active theater and use its presence as a strategic shield, seeking to deter strikes by raising the prospect of radiological release and its attendant humanitarian, environmental, and political costs to complicate the adversary’s targeting calculus. This creates a safety‑deterrence paradox—the reactor might exert strategic leverage while remaining physically vulnerable. But how the adversary weighs those costs cannot be determined.
Conventional plants have demonstrated that nuclear infrastructure can become entangled in conflict, and that the prospect of radiological consequences does not guarantee restraint. Drone strikes around Zaporizhzhia, Chernobyl, and Kursk show how radiological risk shapes targeting, restraint (or the lack thereof), and rhetoric. While these experiences of gigawatt‑scale plants in conflict offer a cautionary precedent, their relevance to microreactors possessing smaller radioactive inventories should not be presumed comparable. The risk nevertheless persists, as the strategic significance also depends on their location, military function, and the potential for operational disruption or escalation, leaving them vulnerable as targets and consequential as deterrents. The strategic shield‑sitting duck paradox therefore represents two sides of the same deployment decision rather than separate outcomes.
The strategic effects of such a deployment could unfold in three ways. First, it may work as a deterrent if the anticipated costs of an attack outweigh its expected military gains. Second, it may prompt a diversion, wherein an adversary modifies their targeting choices to avoid the reactor while continuing military operations. Or third, it may contribute to escalation if the adversary proceeds despite these risks, intensifying the conflict’s political and strategic consequences.
Military programs pursuing mobile nuclear power for energy resilience, of course, do not directly establish an intent to use reactors as strategic shields to deter adversary attacks. The strategic shield concept is best understood as a risk to examine before deployment practices become established.
Governing the Next Nuclear Frontier
One possible governance approach would be to expand the existing regional non‑attack principles to include parallel commitments around the placement and protection of military assets while simultaneously extending such principles to a broader global framework.
Instead of solely relying on designating nuclear facilities as protected sites, states should consider a multilateral moratorium on deploying microreactors at military installations until their strategic and conflict‑related risks have been independently assessed. However, given likely political objections, a more feasible starting point would be a time‑bound pause on operational deployment in conflict‑prone environments, pending consensus on common safeguards. This would preserve the option of military applications while establishing conditions for responsible deployment before practices become entrenched.
The mobility aspect also requires distinct confidence‑building measures. The International Atomic Energy Agency should develop a model‑notification framework drawing on existing nuclear risk‑reduction arrangements. It could also establish dedicated working groups to examine extreme and conflict‑related scenarios involving small and advanced reactors and other emerging reactor technologies. Building on the IAEA’s existing work on safety, security, and safeguards, the groups could conduct structured scenario exercises and wargame how different reactor configurations might interact with armed conflict, including attacks on or near nuclear facilities, co‑location with military assets, transportable reactors, prolonged loss of external support, and other circumstances that may fall outside conventional design‑basis assumptions.
As reactors move into more varied, dispersed, and specialized settings, the rules governing their protection and conduct in crises will need to be reconsidered accordingly.
Hely Desai is a nuclear futures fellow with the Ploughshares Fund and a visiting research fellow at the Henry L. Stimson Center. She holds a Masters in Philosophy from the University of Cambridge. Her work focuses on nuclear deterrence, defense and security, and the political risks surrounding civilian nuclear energy in Asia.
Image: SSgt Monique Bright via U.S. Air Force