The Three-Body Problem

The quiet start of diplomatic dialogue between Russia and the United States has brought about a new phase in discussions on nuclear arms limitation. While these issues are less urgent than resolving the Ukraine crisis, progress here may be just as challenging. Simply returning to the status quo is not a solution, as the accumulated problems in this sphere are not related to recent tensions.

If the U.S. chooses to address the challenge of two potential adversaries not by expanding its arsenal but by strengthening the resilience of its strategic nuclear forces, it would be a preferable outcome for Russia, as it would not result in a threatening U.S. superiority in offensive capabilities. Naturally, this approach is only viable as long as Moscow remains confident in the ability of Russia’s strategic nuclear forces to deliver a retaliatory strike strong enough to ensure deterrence. This will require continued efforts to enhance the resilience of strategic nuclear forces, including by shielding certain regions with missile and space defense systems.

If Moscow and Beijing decide they can afford to show understanding for U.S. efforts to ensure its own security, this step must be reciprocated. Washington should not insist on drastic reductions in Russia’s and China’s tactical nuclear arms (essentially weapons with “intra-continental” range), as these are not a direct threat to the U.S. Tactical nuclear weapons are a proportionate means of counterbalancing NATO (which includes nuclear powers: the United Kingdom and France), along with South Korea and Japan (threshold nuclear states), and forward-deployed U.S. forces. Moreover, given this practice of forward deployment, U.S. delivery vehicles with “intra-continental” range pose an immediate and strategic threat. Ideally, the U.S. should limit them to current capabilities (such as gravity bombs) or to very small numbers of new types (if developed, for instance, under the submarine-launched cruise missile (SLCM-N) program) so that they serve only as political tools for upholding allied commitments as part of “extended deterrence” or as instruments for isolated, symmetric signaling strikes.

Solving the problem of balancing on a three-legged stool—especially one with uneven legs—is no easy thing, but there is hope if the sides engage in active dialogue, as Russia and China are already doing. This will likely mean abandoning the Cold War’s (imperfectly) symmetrical U.S.–Soviet model in favor of an approach that reflects the diverse needs, capabilities and limitations of all actors.

The quiet start of diplomatic dialogue between Russia and the United States has brought about a new phase in discussions on nuclear arms limitation. While these issues are less urgent than resolving the Ukraine crisis, progress here may be just as challenging. Simply returning to the status quo is not a solution, as the accumulated problems in this sphere are not related to recent tensions.

Media reports that express hope for positive outcomes from the ongoing Russia–U.S. dialogue often raise one issue: the looming expiration of the New Strategic Arms Reduction Treaty, or New START, signed in Prague in 2010 [1]. During Donald Trump’s first presidential term, Russia failed to negotiate with the U.S. either a replacement agreement or an extension of the existing one. That extension was eventually achieved under the Joe Biden administration in early February 2021. As permitted by the treaty, New START was extended once, for five years, and will now expire on February 5, 2026—less than a year from now.

However, for over two years—since February 28, 2023—the accord has been suspended at Russia’s initiative. The main reason cited is that allowing U.S. inspectors to resume visits to Russian strategic aviation bases would be inadmissible, since those same facilities have been attacked by Ukrainian drones, apparently with U.S. intelligence support (mutual inspections had already stopped in early 2020 due to the COVID-19 pandemic). Moreover, while demanding the resumption of inspections at Russian sites, the U.S. could not guarantee timely and unhindered access to its own facilities (as direct flights are suspended). A less publicized but practically important factor was that the agreement required both sides to constantly exchange information about the status and location of strategic delivery vehicles (normally, several messages were sent daily via established communication channels). With Russia taking steps to disperse its bombers involved in the special military operation, this practice was likely seen as unacceptable.

In fact, the suspension amounted to halting the fulfillment of obligations concerning information—Russia decided to continue complying with the treaty’s core provisions on maintaining ceilings for deployed warheads and weapons carrying them, which both countries had reached back in 2017–2018 (since then, no actual reductions in strategic arms have taken place, and the agreement came to serve as a cap, ensuring that the numbers do not grow). Even though mutual suspicions have mounted over the five years without inspections, neither country could significantly exceed the New START thresholds—only partial rearming of ballistic missiles with extra warheads may slip under the radar, but that would not give much of an advantage and, if discovered, would expose the initiator as an aggressor, giving the other side a free hand.

It is fair to assume that Moscow may be ready to resume implementation of the accord if meaningful progress is made in resolving the Ukraine conflict and the most pressing security concerns are addressed. However, this raises the question: what is the point? Beyond symbolism, of course—it would allow New START to end on a positive note, for example, by presenting figures on warheads and their delivery vehicles on February 5, 2026, which confirm both sides remained within the treaty’s limits up to its final day. And they would claim to have acted purely out of “responsibility” and “sincere commitment to nuclear disarmament” (at least, it would be a positive message to non-nuclear states at the 2026 Review Conference of the Parties to the Treaty on the Non-Proliferation of Nuclear Weapons).

New START does not necessarily have to end next February: although the treaty does not provide for more than one extension of up to five years, neither does it mention any “suspension.” This is a bilateral accord between sovereign actors, and nobody can stop the countries’ leaders from signing a stopgap agreement consisting of a few clauses that say, “New START shall remain in force until…” (for example, extending the treaty by the amount of time it was suspended so it can “serve out its full term”).

There is already a relevant precedent—following the failed ratification of START II, the Strategic Offensive Reductions Treaty (SORT) was quickly signed. It did not replace START I but extended it, setting new quantitative limits. This approach used to seem impractical, since extending a treaty beyond its original framework would require ratification by the U.S. Congress, which has historically been reluctant to support arms control agreements. But the political landscape on Capitol Hill today is unique: most Republicans are on board with the president’s agenda, and Democrats, for ideological reasons, tend to be less critical of arms control, which means Trump could theoretically push a New START 2.0 through Congress. The bigger issue is that neither side is likely interested in it anymore, as New START is falling out of step with the evolving balance of power.

Brave new world

Ivan Timofeev:
Nuclear Disarmament. The End of History?

In recent years, the concept of a “third nuclear age” has gained traction in Western discourse to describe the current period in history. It follows the Cold War between the U.S. and the Soviet Union and the era of frequent and deep reductions in nuclear arsenals by major powers in the 1990s and 2000s. A defining characteristic of the new phase, according to American experts, is the growth of China’s nuclear capabilities and the emergence of new nuclear powers (for obvious reasons, with a focus on North Korea developing effective nuclear forces and Iran’s threshold status). As Moscow and Washington cut their nuclear arsenals, NATO allies came to play a more prominent role—though they too reduced their stockpiles, albeit not as drastically.

U.S. officials’ estimates of China’s nuclear arsenal and its projected growth have been rising year by year—this trend was seen under the previous administration and is expected to continue, at the very least, under the current one. According to the latest official Pentagon assessment, China had more than 600 operational nuclear warheads as of mid-2024 and will have over 1,000 by 2030. In this context, for several years now, the U.S. military has pointed to the challenge of ensuring “simultaneous but independent deterrence of two close adversaries.”

The wording suggests that the problem in U.S. military planning in this sphere lies in the uncertainty over how close Russia and China will be in the future and whether the U.S. strategic nuclear forces should be tasked with deterring the combined nuclear arsenals of Russia and China. In the most straightforward way, this requires if not strict parity with the aggregate capabilities of the two, then at least something close to that level. At the same time, the process of developing such capabilities by the U.S. will prompt Moscow and Beijing to take reciprocal measures, including closer cooperation—creating a kind of self-fulfilling prophecy. This, presumably, is why the idea of “independent” deterrence of the two adversaries is emphasized.

From the Russian side, American complaints about a “unique and unprecedented challenge” are often met with irony, given that in the 1970s and 1980s, it was the U.S. that sought to engineer a similar problem for the Soviet Union by building an informal bloc with China. While there are parallels—Moscow did face the pressing problem of ensuring deterrence on its eastern flank—the situations were far from identical. At the time, the Soviet Union’s main concern with China had more to do with conventional conflict. Beijing’s nuclear arsenal was small, technically unsophisticated and far inferior to those of the United Kingdom and France (whose combined nuclear capabilities, in turn, were negligible compared to the Soviet and U.S. stockpiles [2]). Due to geographic proximity, the Soviet Union could rely on different means to deter China than it used against the U.S. (for example, about one-third of the Pioneer intermediate-range missile systems were deployed east of the Urals). Today, however, the U.S. faces the challenge of maintaining nuclear deterrence of one peer adversary and the other near-peer (with a trend toward narrowing the gap). Both are located at equal intercontinental distances, which makes it difficult to allocate forces to deter only one of them in a way that would not lead to a security dilemma for the other [3].

At this point (if not earlier), readers may wonder why U.S. strategic concerns should matter to us. After all, they are of Washington’s own making, and it is tempting to simply sit back and enjoy watching the U.S. struggle. The problem is, the most straightforward and simple solution it might choose is a significant expansion of nuclear capabilities over the coming decades—aiming not necessarily to match the combined Russian and Chinese arsenals, but at least to far surpass each individually.

Such ideas have long been floated by the most hawkish think tanks, and while this would not be important in itself, they now appear to have been adopted by Pentagon officials. For example, at a conference in March 2025, Gen. Anthony J. Cotton, chief of U.S. Strategic Command (USSTRATCOM, the unified command responsible for all components of the U.S. nuclear triad operated by the Air Force and Navy), said the military should acquire at least 145 new B-21 Raider strategic bombers (up from “a minimum of 100” officially envisioned), ramp up production of AGM-181 LRSO nuclear-armed cruise missiles to equip both the B-21 and the upgraded B-52 bombers, and expand the fleet of Columbia-class ballistic missile submarines beyond the current plan of 12 to replace 14 Ohio-class submarines. He stopped short of proposing an increase in the number of new land-based intercontinental ballistic missiles (ICBM) LGM-35 Sentinel—most likely because the program has been getting bad press after a spike in cost projections made headlines. Senior U.S. military officials had previously avoided talk of expanding the nuclear stockpile, likely because all efforts were focused on safeguarding funding for the baseline levels. It now seems they have sensed that political winds have changed behind closed doors in Congress and the Pentagon, and it is possible to push for more.

How will Russia and China respond to a major build-up of U.S. strategic nuclear forces?

Aleksey Arbatov:
Nuclear Doctrine and Strategic Stability

Strict numerical parity is not essential for credible nuclear deterrence. Russia has not even maxed out its New START limits, often falling well below them (especially in terms of deployed delivery vehicles, which are the most expensive to produce and maintain). But given that the U.S. has nuclear-armed allies, a substantial arsenal of strategic-range non-nuclear weapons, which is growing both quantitatively and qualitatively (and can also be used for counterforce strikes against strategic nuclear forces and their command-and-control systems), and has entered a new phase in building global missile defense, Russia’s military and political leadership will not accept a big U.S. advantage in the number of strategic nuclear weapons. A quantitative and qualitative advantage in tactical nuclear arms with “intra-continental” range will not be of much help to Russia in this situation, as it can only partially offset NATO’s weight. At the same time, Moscow does not perceive its relations with Beijing as a firm military alliance (and it is hard to imagine that it will in the future), meaning that the authorities cannot rely on China’s strategic nuclear forces to strengthen Russia’s own deterrence.

Because Russia kept its nuclear weapons complex going through the 1990s and then developed it further, the country is surely capable of expanding its strategic nuclear forces at a pace comparable to that of the U.S. (where the industry collapsed, delaying the resumption of large-scale thermonuclear warhead production until the mid-2030s). However, this growth will require additional spending for both acquisition and maintenance, affecting state and defense budgets alike. It is important to remember that this is not a guns-versus-butter dilemma: when military spending is always capped, every extra missile in a silo comes at the cost of other weapons that are not procured and personnel who are not assigned to other branches of the armed forces. Yet these extra squadrons of aircraft, tank regiments, drones and other assets are far more likely to see real combat, and success in conventional warfare can influence whether escalation to higher levels of conflict becomes necessary.

From Russia’s perspective, a scenario where the U.S. significantly expands its strategic arsenal is undesirable and should be avoided, including by proposing alternative solutions to Washington. Given today’s increasingly tense global environment, further nuclear arms reductions are likely off the table for now. All the more so because Russia is already uneasy about making deeper cuts unless the UK and France join the effort. And the prospects for a global “nuclear agreement” (among at least the P5, with China potentially pushing to include India, which, in turn, may insist on Pakistan’s involvement) in the next decade are quite bleak.

Amid its escalating standoff with the U.S. over the past decades, China’s military and political leadership clearly aims to build strategic nuclear forces that ensure not minimal, but credible and guaranteed deterrence. This does not necessarily require achieving strict parity (especially considering that the U.S. estimates of 600… 1,000… 1,500… Chinese warheads also include those deployed on regional-range delivery vehicles), but a tenfold difference in the number of warheads capable of reaching the adversary’s mainland is more unacceptable. Moreover, China is surrounded by several U.S. allies and American forward bases, while many other countries in the region have varying degrees of hostility toward it (as the confrontation between the U.S. and China deepens, the number of neutral states will decline, and so far, the shift has been to Beijing’s disadvantage). Probably, China can mass-produce missiles, but its warhead production is lagging behind because the industry was built around the logic of minimal deterrence. Since the U.S. has a major lead, its move to build up the arsenal would be even more problematic for Beijing than for Moscow.

Balancing on a three-legged stool

Ivan Timofeev:
Trump and Gorbachev: Will the US Suffer the Fate of the USSR?

Resolving the dilemma of strategic arms limitations in the current historical period is a formidable task—no less, and perhaps even more difficult than it was between the U.S. and the Soviet Union, where negotiating the rules of the game took up at least half of the Cold War era. Asymmetry is seen as a challenge and a core weakness in the U.S.–Russia–China triangle, but rather than trying to overcome it by juggling numerical limits (for example, by trying to revisit the 1922 Washington Naval Treaty), perhaps it should be used as a tool for addressing the problem.

China’s growing strategic nuclear forces have been a hot topic for years—the previous U.S. administration even pushed this issue in negotiations with Russia about extending New START. But this year, a new “troublemaker” has entered the scene: the Golden Dome, a next-generation U.S. national missile defense system with space-based strike capabilities, which was proposed by Trump after his return to the White House [4]. The executive order signed on January 27, 2025, outlines Washington’s “peace through strength” policy as focused on defending citizens and critical infrastructure against missile and aerial attacks, as well as guaranteeing secure second-strike capability.

To achieve this, the plan proposes deploying space-based sensor systems, boost-phase interceptors stationed in orbit, as well as underlayer and terminal-phase intercept capabilities (likely referring to anti-ballistic missiles). The executive order also calls for exploring “non-kinetic capabilities to augment the kinetic defeat of ballistic, hypersonic, advanced cruise missiles, and other next-generation aerial attacks,” which can be interpreted both as a green light to continue past experiments with lasers and as an indication that anti-ballistic missiles (now including those mounted on satellites) remain the primary means of interception at this stage. The main difference between the proposed Golden Dome and the current U.S. national missile defense system is the stated objective of defending against not only “rogue states” but also peer adversaries and defeating a “countervalue attack” [5].

The architecture of the Golden Dome remains unclear three months after its announcement. In fact, the concept appears hazy even to its proponent: the Pentagon is still soliciting proposals from the industry and is unlikely to finalize the design of the future system in less than a year. Yet leaked industry proposals raise doubts as to whether Trump’s bold goal is primarily a propaganda move or something the administration actually intends to pursue.

In late March, the Booz Allen Hamilton company unveiled a concept for a constellation of “up to 2,000” satellites, dubbed Brilliant Swarms, which would double as launch detection sensors and interceptors. They are supposed to be put in polar orbits at 300 to 600 kilometers altitude. Each satellite would weigh between 40 and 80 kilograms, making the constellation quick to deploy. The idea and even its name seem like a nod to the Brilliant Pebbles project, a central element of the Strategic Defense Initiative (SDI), which was selected in the late 1980s as one of the most feasible for implementation. Still, the credibility of Booz Allen Hamilton’s proposal is debatable—it seems to borrow heavily from the original SDI concept, dressed up with modern buzzwords such as AI, machine learning, interconnected satellite mega-constellations, and more.

The original “pebble” had roughly the same mass, but about 90% of it was fuel (basically, the Brilliant Pebbles interceptor satellite was a compact missile with infrared homing, enclosed in a container with a small solar panel). A satellite in the modern “swarm” would need to carry inter-constellation communication systems, space situational awareness sensors (as the developers claim anti-satellite capabilities), missile detection and targeting tools, and an onboard computer. These components reduce the share of the satellite’s mass allocated to engines and fuel, which, in turn, limits the acceleration it can achieve.

The slogan “AI will allow selecting the best-positioned satellite for attack” looks like something awkwardly generated by artificial intelligence itself. Such a constellation would not have much difficulty in deciding which satellite to activate: every unit with even a slim chance of success should try to intercept an ICBM. While “2,000 satellites in polar orbit” may sound impressive, in practice they would be evenly spread out around the globe. According to computational models, only four to seven satellites would be over China at any given moment. In a scenario where launches are even barely back-to-back, there would be no time for choice—not the interceptor’s onboard intelligence, but its dynamic features would matter. But this is hardly surprising, given that the authors are not engineers but consultants and cybersecurity experts.

Much less is known about proposals from companies specializing in space technology. Sierra Nevada, for example, reportedly proposed larger orbital stations that function as “magazines” loaded with multiple interceptors, which could be supported and supplied by the Dream Chaser—a shuttle in development by the company. Boeing is planning to involve its X-37B military mini-shuttles in testing various parts of the Golden Dome and engage its subsidiaries in the development of satellites that can detect and track missiles.

Dmitry Stefanovich:
On the Enduring Value of Nuclear Non-Proliferation

For obvious reasons, media reports say SpaceX’s proposal has emerged as a frontrunner. Leaked details suggest that Elon Musk’s company pitched a plan to deploy 400 to over 1,000 satellites of its own design that would sense missiles and track their movement (likely to be based on the Starshield platform, a derivative of Starlink communication satellites). In addition, a separate fleet of 200 attack satellites is expected to be built by another company. As there would be so few of them, each unit would clearly be equipped with multiple interceptors intended to shoot down post-boost vehicles or warheads after the missile boost phase. Even so, the dispersion of the interceptor constellation around the globe means the system would not counter anything more than a minor strike.

And if the system’s design reflects the client’s requirements (which are presumably well known to Musk), this invites several interpretations. The key one is—aware of the immense cost, complexity and potential risks (in terms of adversaries’ response), the U.S. is once again producing a system with limited capabilities, even while attempting to build a comprehensive global missile defense shield. This time, however, it also seems relevant in a confrontation with major nuclear powers.

The system is not designed to provide guaranteed defense against a large coordinated nuclear missile attack—salvo launches from position areas would overwhelm the satellites’ defenses. The space shield could also be breached using special means, which were under discussion back in the 1980s and range from ground-based laser weapons, similar to Russia’s Peresvet, to high-altitude nuclear explosions. Now that the U.S. is moving to reintroduce the concepts of that era with modern technology, Russia (and China, which has undoubtedly researched this issue) may also revisit these old ideas. Naturally, there is a risk to the effectiveness of a “deep” second strike (one launched after enemy warheads have hit their targets using surviving assets), but Moscow and Beijing are drifting in their nuclear doctrines toward a launch-on-warning strategy (launching a strike while enemy missiles are still in the air). This shift is driven less by missile defense systems and more by advancements in non-nuclear high-precision weapons, including “hypersonic” systems.

The Golden Dome, in its full configuration based on current technology and with consideration to capabilities for missile defense penetration that major missile powers will develop, would be able to intercept only a portion of a missile salvo launched by a “peer or near-peer” adversary. This is where the system’s stated objective starts to sound disingenuous: this approach is almost meaningless for defending against countervalue attacks. It hardly matters for a major urban area whether it gets hit by five warheads or fifteen. Given that fires, destruction and the collapse of medical services for the injured would be the main problems, the number of casualties would not be proportional to the total megatonnage.

When it comes to defending against countervalue attacks, where each small, protected target must be hit with high precision, adding a “reduction factor” dramatically increases the required number of delivery vehicles to such an extent that the attack becomes unfeasible during planning. So the actual purpose of the Golden Dome may not be the “defense of citizens” but guaranteeing “secure second-strike capability.” As decades before, this guarantee would underpin mutual security of superpowers. The U.S. would also feel less vulnerable to new nuclear-armed states, though not so invincible as to risk reckless actions: even a single missile breaching defenses in a conflict with North Korea would be unacceptable.

***

If the U.S. chooses to address the challenge of two potential adversaries not by expanding its arsenal but by strengthening the resilience of its strategic nuclear forces [6], it would be a preferable outcome for Russia, as it would not result in a threatening U.S. superiority in offensive capabilities. Naturally, this approach is only viable as long as Moscow remains confident in the ability of Russia’s strategic nuclear forces to deliver a retaliatory strike strong enough to ensure deterrence. This will require continued efforts to enhance the resilience of strategic nuclear forces, including by shielding certain regions with missile and space defense systems.

If Moscow and Beijing decide they can afford to show understanding for U.S. efforts to ensure its own security, this step must be reciprocated. Washington should not insist on drastic reductions in Russia’s and China’s tactical nuclear arms (essentially weapons with “intra-continental” range), as these are not a direct threat to the U.S. Tactical nuclear weapons are a proportionate means of counterbalancing NATO (which includes nuclear powers: the UK and France), along with South Korea and Japan (threshold nuclear states), and forward-deployed U.S. forces [7]. Moreover, given this practice of forward deployment, U.S. delivery vehicles with “intra-continental” range pose an immediate and strategic threat. Ideally, the U.S. should limit them to current capabilities (such as gravity bombs) or to very small numbers of new types (if developed, for instance, under the submarine-launched cruise missile (SLCM-N) program) so that they serve only as political tools for upholding allied commitments as part of “extended deterrence” or as instruments for isolated, symmetric signaling strikes.

Solving the problem of balancing on a three-legged stool—especially one with uneven legs—is no easy thing, but there is hope if the sides engage in active dialogue, as Russia and China are already doing. This will likely mean abandoning the Cold War’s (imperfectly) symmetrical U.S.–Soviet model in favor of an approach that reflects the diverse needs, capabilities and limitations of all actors.

1. START I was an agreement signed in 1991, START II was signed in 1993 but never entered into force, while its follow-up was supposed to be START III, which was discussed in the second half of the 1990s.

2. To put things into perspective, the Soviet Union had about 150 to 200 times more nuclear warheads than China in the 1980s.

3. It is worth recalling the Soviet missile system Pioneer, which was mentioned earlier. When it was deployed, the U.S. reacted calmly, while the “panic” that ultimately led to the Euromissile crisis was stirred up by NATO’s European members.

4. The U.S. program was initially dubbed Iron Dome for America—evidently inspired by the success of Israel’s missile defense during a limited exchange of fire with Iran in late 2024—but was renamed a month later. The proposed space-based strategic missile defense naturally has nothing in common with Israel’s short-range tactical system, which is designed to intercept unguided rockets. See: Golden Dome replaces Iron Dome: Pentagon renames missile defense initiative // SpaceNews, 28.02.2025. https://spacenews.com/golden-dome-replaces-iron-dome-pentagon-renames-missile-defense-initiative/

5. Two widely used terms in nuclear strategy are countervalue and counterforce strikes. The former target key infrastructure and economic facilities (and other euphemisms developed over the past half-century to avoid explicitly saying “cities”), while the latter are aimed at nuclear forces and their command-and-control systems.

6. The Trump administration showed interest in this issue during its previous term, as the development of mobile land-based systems in the U.S. was believed to be possible if the security of subsurface launch platforms was compromised by advances in unmanned underwater vehicles.

7. In the context of the “strategic partnership” between Russia and China, the topic is not typically discussed, but this includes mutual deterrence capabilities in the hypothetical event of a return to the tensions of the 1960s–1980s. Military planning in this area, which looks decades ahead, must take into account a range of possible scenarios.