11 minutes
The world is more than three quarters of the way through the first century of living with nuclear weapons. Despite crises and close calls, this period has been an era of nuclear peace, with no nuclear detonations in anger since August 1945. At the end of the Cold War, the salience of nuclear weapons receded but now they once again cast a long shadow over world affairs. From conflict with Iran over its nuclear program, to serious concerns of Russian nuclear employment in the fall of 2022, to the transformation of China’s nuclear force, nuclear weapons are ubiquitous. At the same time, new technologies and approaches to nuclear power offer opportunities for nearly unlimited clean energy even as other technologies like artificial intelligence promise transformation as profound as the Industrial Revolution.
For scholars and policymakers alike, looking ahead to the impending second nuclear century requires reflecting on three distinct but related arenas. The first is the history of the Cold War and nuclear weapons, about which more information has become available in recent years. The second is the 21st-century evolution of the nuclear order built on the strategic choices of the United States and other nuclear powers, especially China. The third is the advance of technology, from fusion energy to artificial intelligence.
The Unknown History of the Cold War Competition
In 2026, the Cold War has been over for nearly as many years as it endured. Yet it remains the only global competition between nuclear armed superpowers and thus is a major source of what data we have on living with nuclear weapons. Moreover, thanks to declassification and other revelations we continue to learn more about the many facets of nuclear competition.
For example, the now declassified 1981 History of the Strategic Arms Competition 1945–1972 runs to over 1,000 pages.[1] A magisterial work drawing on a wide array of classified and unclassified sources available at the time, it is both invaluable and incomplete. The history is detailed but lacks the information that has become available from Soviet archives as well as the final years of the Cold War, which includes the Carter-Reagan nuclear modernization. Recent scholarship and other declassified works continue to fill out our understanding of the dynamics of the nuclear competition between the United States and Soviet Union.[2] Yet we still know relatively little about the reasons the Cold War competition went the way it did.
If one had predicted in 1959 following 1957’s Sputnik launch that in only 20 years the United States would thoroughly dominate the military and intelligence uses of space, one would have been greeted with skepticism.[3] However, by 1980, this was the reality as the Soviet Union lagged far behind in satellites for early warning and intelligence collection.[4] Was it because the Soviet Union simply had fewer resources that it was not in fact a superpower capable of internally balancing against the United States?[5] Possibly, yet on paper the Soviet Union had access to massive amounts of every natural resource as well as brilliant scientists and engineers. Was it perhaps the inferiority of the Soviet political-economic system?[6] Again plausible, but the Soviets excelled in some aspects of the broader competition. In 1962, the Soviet strategic nuclear force was small and woefully inadequate in terms of survivability. Yet by the end of the decade the Soviet strategic force had reached rough quantitative parity with the United States, with substantial increases in survivability.
Why then did the Soviet Union have so much less success in the space competition, given its importance to the nuclear balance as well as international prestige? Similar questions arise about the undersea competition, where United States dominance only began to erode modestly at the very end of the Cold War.[7] Allies surely played a part, from Tromsø to Pine Gap, in US dominance under the sea and above the atmosphere, but how and to what extent are underexplored, particularly from a theoretical perspective.[8] Explanations at the level of the individual firms and organizations may be important as well, but virtually nothing comparative has been done here.
These questions are of more than simple historical interest. If the second nuclear century will be marked by competition, then understanding the determinants of the last such competition are crucial to looking ahead to the future. Simply asserting the futility of competition is an insufficient guide, given that such competition was hardly futile in the Cold War.[9]
Choices About Strategy and Order in the Second Nuclear Century
The Cold War produced the principles for nuclear order that governed much of the first nuclear century (and I use the term “order” here rather loosely). The United States and Soviet Union sought to limit the proliferation of nuclear weapons, with some success, particularly in the key case of Germany, and to avoid nuclear war except as (perhaps) the last resort. The result was nuclear order with two dominant nuclear powers and a handful of much smaller nuclear states, with tacit and explicit agreements about the status quo. This order proved stable and durable, built on both liberal internationalist institutions (e.g., the International Atomic Energy Agency) and realist hard power (US alliances and extended nuclear deterrence), with the Nuclear Nonproliferation Treaty at its center.[10] US-Soviet bilateral arms control further cemented this order by enabling the shaping and management of the military competition described in the previous section.[11]
The principles and foundations of this nuclear order are now in flux, with some abandoned already and some under question. Most seismically, China, long held out as a paragon of minimal nuclear forces being sufficient for deterrence, has totally transformed its nuclear forces and posture.[12] Within a decade it will join Russia as a nuclear peer of the United States, resulting in the so-called “two peer” or “three body” problem.[13]
Beyond China, Russian conventional weakness, on full display in Ukraine, will likely reinforce its reliance on nuclear weapons to achieve its aims.[14] This strategy was most clearly underscored in the fall of 2022, when the United States and its allies believed the possibility of Russian nuclear employment to forestall conventional catastrophe was close to a coin toss.[15] Such nuclear use would have brought to a close the era of nuclear non-use, with unknowable consequences—an outcome avoided in 2022 but perhaps not in the next such crisis. Russian conventional weakness and increasing reliance on nuclear weapons, potentially including weapons in orbit, makes future crises more fraught than those of the old nuclear order, where Soviet conventional strength was taken as a given. Moreover, the demise of bilateral arms control, with the end of the New START agreement in February 2026, removed another foundation of the old nuclear order.
The nuclear risks of a transformed China and conventionally enfeebled Russia have clear implications for US allies dependent on extended nuclear deterrence. Yet the commitment of the United States to extended nuclear deterrence is viewed as shakier than ever, particularly by European allies.[16] Such changes in turn lead to questions about nonproliferation and how the United States (along with China and Russia) would respond to so-called “friendly” proliferation by US allies.[17] This is in addition to ongoing concerns about proliferation in Iran, the growing North Korean nuclear arsenal, and the uneasy nuclear relationship between India and Pakistan. The latter saw one of the most serious direct clashes between nuclear-armed states in 2025—and the United States played a key role in mediating an end to the crisis.[18] These developments—the evolution in the extended deterrence landscape, concern about both allied and other proliferation, and the growing possibility of crises between nuclear states—all underscore the enduring importance of US power to the nuclear order.
Yet even if the United States wants to bolster extended deterrence and ensure its power continues to underpin the nuclear order, there are questions about its ability to do so in the “two peer” environment. A military competition with China and Russia in the 21st century would look different than the competition with the Soviets, and the United States might not prevail. Yet the United States retains strong advantages over both Russia and China, though not the same advantages over both, as I have argued elsewhere. In brief, Russia has operational experience and a large nuclear complex but few resources. China has vast (though not unlimited resources) but limited operational experience and a small (but growing) nuclear complex.[19]
However, without understanding what was consequential during the Cold War we cannot begin to predict what is important now and what will be in the future. Regardless, a second nuclear century with less (or no) US extended deterrence and more nuclear proliferation by US allies would be radically different than what has gone before, with potentially far-reaching consequences from more nuclear accidents in nascent proliferators to perhaps a more heavily armed but deeper nuclear peace.[20] As with the Cold War history, the consequences of these choices means they should merit the highest level of attention from scholars and policymakers.
Technology in the Second Nuclear Century
Beyond choices about strategy and order, the second nuclear century will be molded by technology that is very different than that of the first.[21] In terms of nuclear energy, the first nuclear century was dominated by early nuclear fission reactor designs. These reactors had a number of limitations, including with safety and enormous initial capital investment. Accidents at Three Mile Island, Chernobyl, and Fukushima underscored these limits and stunted the development of nuclear power that once promised energy “too cheap to meter.”
As we approach the second nuclear century, developments in both fission and fusion energy may, after false starts, yield a renaissance in nuclear power. Small modular fission reactors, along with other alternative reactor designs, offer more efficiency, greater safety, or reduced capital costs—perhaps even all of the above. Fusion power offers potentially limitless power with minimal radioactive waste and no need for fissile material like uranium or plutonium.
Yet these advanced power sources also yield different challenges for proliferation than older systems. Fusion power techniques are not easily weaponizable but will, if widely adopted, expand the knowledge base of nuclear fusion processes and the ability to experiment with them—potentially highly useful for proliferators seeking a thermonuclear weapon. Right now, the United States has invested massively in just two sites—the National Ignition Facility at Lawrence Livermore National Laboratory and the Z-machine at Sandia National Laboratory—to conduct research on fusion relevant to nuclear weapons. In the future, potential proliferators may be able to turn to a number of commercial technologies or providers to enable their research. Even if fusion power never becomes commercially viable the vast private sector investments being made now will still create a pool of expertise and equipment potentially relevant to thermonuclear proliferation—a pool with perhaps no economically lucrative future.
Small modular reactors will potentially enable cheap and stable power in remote locations. But inherent mobility and the dispersal to remote locations makes them much more challenging to monitor than the large reactors of today. Moreover, many of these new reactor designs use high-assay low enriched uranium (HALEU)—a form of uranium that is well short of “weapons grade” but much closer to being useful in a weapon than fuels typically used in commercial reactors. This will mean more uranium enrichment capacity and technology around the world—another proliferation risk in addition to the reactors themselves. Safeguards for these new systems are possible but will require different approaches than today, and thoughtfulness now on the best ways to proceed.
Beyond nuclear power, artificial intelligence promises profound military and economic change. Many of these changes will impact nuclear weapons, from design to survivability. Ukraine’s Operation Spiderweb—a clandestine attack on Russian strategic bombers—used partly autonomous drones.[22] A fully autonomous version of the attack is probably already possible and may become even easier in the future. The confluence of artificial intelligence and nuclear command and control raises the specter of films like WarGames and The Terminator becoming reality—an area attracting much study but with much more granular work to be done.[23]
As we approach la fin de siècle nucléaire, the good news for scholars and policymakers is both that we survived our first nuclear century and there are many interesting questions about our nuclear past, present, and future that remain. The bad news is there are many interesting questions about our nuclear past, present, and future that remain—and survival may depend on the answers. The next nuclear century awaits.
