Table of Contents


The 1967 Harmel Report introduced NATO’s dual-track policy of maintaining deterrence while promoting détente: the policy of easing hostilities, including through arms control.1 The report argued for balanced force reductions in Europe, which defined NATO’s approach to disarmament and arms control for the next several decades. That approach resulted in a structure of nuclear and conventional arms control regimes that stood the test of time.

Unfortunately, in recent years, this has no longer been the case. Russia’s violations of the Intermediate-Range Nuclear Forces (INF) Treaty and other regimes have cast a shadow over the future of arms control. Technological developments, such as unmanned systems, robotics, and cyber capabilities, also raise questions about the applicability of the traditional approach to arms control treaties—that is, a focus on reducing the numbers of weapons systems. The rise of China as a world power has introduced a new challenge, in particular to nuclear arms control.

Dick Zandee
Dick Zandee is head of the Security Unit and a senior research fellow at the Clingendael Institute in The Hague, the Netherlands.

NATO is adapting its deterrence posture in response to these changes. The four NATO battle groups in the Baltic states and Poland are an answer to the increased threat from Russia. The allies are also boosting high-end fighting capabilities and reinforcing their defenses against cyber challenges.

The alliance’s arms control and disarmament policy needs to undergo a similar update. The allies should consider how they can overhaul their general approach in this area.

Issues at Stake

Challenges of Arms Control

For the alliance, the immediate challenge in arms control lies in the breakdown of existing treaties and agreements. In 2002, the United States withdrew from the Anti-Ballistic Missile (ABM) Treaty, which limited the number of strategic missile interceptors that the United States and Russia could possess. In 2007, Russia suspended its participation in the Conventional Forces in Europe (CFE) Treaty. Moscow is also hampering information exchange and only partly implements other confidence-building measures contained in the Vienna Document of the Organization for Security and Co-operation in Europe (OSCE) and the Treaty on Open Skies, which establishes a program of unarmed surveillance flights over the territory of participating states. This has increased uncertainty and the risk of an escalation in times of crisis.2

Noncompliance is one issue; the relevance of the present regimes is another. They were created in a different time, amid a different political and military threat: the bloc-to-bloc confrontation of the Cold War.3 The United States rightly pointed to changed circumstances when it withdrew from the ABM Treaty. However, new tensions and open war have since returned to the European mainland. The numbers of Russian provocations and violations of allied airspace, in particular in the Baltic, are increasing.4 The need for renewed arms control and other stability measures is back and more pressing than ever. The question is: What sort of arms control will work in this new security environment and against twenty-first-century military technology?

With regard to nuclear arms, the only existing treaty that bans a complete category of nuclear weapons—the INF Treaty—is dead, having expired on August 2, 2019. The New Strategic Arms Reduction Treaty (START) could be next. In force since 2011, it needs to be extended before February 2021, and the two signatories—the United States and Russia—have not said whether it will be extended, amended, or suspended.

Whatever New START’s future may be, the whole nuclear arms control regime between Washington and Moscow is currently under pressure and could collapse. A breakdown would remove any limits on the two sides’ nuclear weapons, in both a quantitative and qualitative sense. Moreover, the existing transparency about the U.S. and Russian strategic nuclear arsenals would also disappear, because information exchange and verification measures would no longer apply. In a worst-case scenario, a new nuclear arms race might follow, while the risk of misjudgment and escalation could rise in a crisis.

Technological developments are increasingly endangering the nuclear arms stability that has characterized the Moscow-Washington relationship for a long time. The introduction of hypervelocity reentry vehicles, low-yield nuclear warheads, faster cruise missiles, more accurate targeting technology, and other innovations may increase the temptation to use nuclear weapons in a crisis. States that possess nuclear weapons might conclude that new technologies allow them to deploy these weapons in such a precise and limited manner as to avoid the risk of escalation into an all-out nuclear exchange.

The same effect may arise from the narrowing gap between heavy conventional blasts and low-yield nuclear detonation. A massive conventional bomb may be considered a nuclear weapon, triggering a nuclear response. Nonmilitary means—cyber attacks against command-and-control centers in particular—can endanger nuclear stability as well, leading to unintended escalation. Naturally, countermeasures will be taken to prevent the compromising of nuclear command-and-control systems, but cyber and artificial intelligence (AI) are developing so fast that the question can be posed whether the usual action-reaction cycle applies.

The rise of China as a new world power creates additional problems for arms control and disarmament treaties and agreements. One of the reasons for the breakdown of the INF Treaty is that it does not cover China’s intermediate-range nuclear weapons, creating imbalances with both the United States and Russia. Other rising powers, such as India, are also expanding their nuclear arsenals. New arms control initiatives should be broader than the traditional bilateral U.S.-Russia context. Yet the more states are involved, the more difficulties will be brought to the negotiating table.

Opportunities and New Approaches to Arms Control

On the one hand, the trends point toward increasing difficulties for old-style arms control. While the United States and Russia might still have a bilateral interest in reducing and controlling numbers of nuclear weapons, China and other nuclear weapons states are reluctant to join any arms-reduction regimes.5 Beijing points at the imbalance of nuclear arsenals: the United States possesses 6,500 nuclear warheads, Russia 6,185, but China only 290.6 From the Chinese perspective, the United States and Russia have to start disarming toward a number of warheads that is closer to China’s before Beijing will join any negotiations.

On the other hand, China and India might be interested in risk-reduction measures, as these would serve their interests as well. Negotiations on nuclear risk-reduction measures could focus on limiting unauthorized use (someone using a nuclear weapon without official authorization), unintended use (for example, by accident or due to a technical error), and intended use based on incorrect assumptions (authorized use that later appears to be based on incorrect information, misunderstandings, or misperceptions). These measures could deploy any combination of new approaches, including:

  • improving training for nuclear emergencies;
  • increasing transparency regarding nuclear capabilities, doctrines, postures, and other related policies;
  • improving communications between nuclear weapons states;
  • de-targeting nuclear weapons;
  • increasing the security of launching systems;
  • de-alerting nuclear weapons;
  • increasing the decision time for nuclear weapon use;
  • raising the threshold for use;
  • eliminating certain types of nuclear weapons; and
  • limiting the numbers and locations of nuclear weapons.

In the realm of conventional forces in Europe, too, the likelihood of negotiating new quantitative arms control agreements with Russia is rather low. Given that some existing nuclear and conventional arms control regimes are probably on their way out, and are less and less relevant, it might be worth exploring the scope for new negotiations focused on risk-prevention and risk-reduction measures.

One idea, launched by the OSCE network, is the new approach to risk-reduction measures for the NATO-Russia contact zone in the Baltic.7 The proposed restrictions on troop deployment would probably favor Russia because of its short supply lines and should therefore be rejected in their current form, but there is ample scope to explore new measures for limiting the size of military activities, restricting snap exercises, and increasing notification and observation provisions. New measures could also focus on preventing and reducing the risk of airspace violations, for example by prohibiting the deactivation of transponders, which makes it impossible to communicate with pilots flying aircraft in contested airspace.


NATO’s arms control policy should continue to be based on the combination of deterrence and détente. Verification also remains an essential element for arms control and disarmament regimes. Yet the alliance must adapt its approach to reflect changed geopolitical realities.

In the past, allies’ arms control policies have been largely based on setting limits on numbers of weapons, either in support of the United States or in the context of conventional forces. That made sense during the Cold War and in the first two decades after it, when the two capabilities that mattered most to military balance were nuclear and conventional weapons.

Today and in the future, new forms of warfare come into play. The weaponization of AI increases the risk of loss of human control over the use of force. A new agreement is needed to address the dangers of conventional weapons systems deployed with humans out of the loop. A new regime that limits the most destabilizing kinds of cyber attacks is also overdue. It should limit or ban those that target nuclear command-and-control and critical civilian nodes or automated cyber attacks controlled by AI.

There is little political will on any side to negotiate new quantitative limits, in particular for conventional weapons. But there is an urgent requirement for new talks on measures to reduce risks and prevent escalation in a crisis. These measures should aim to prevent the misuse of nuclear weapons and ensure that the nuclear threshold is raised, not lowered. To this end, the alliance should give priority to considering the prohibition of low-yield nuclear weapons. In the conventional area, the same approach should prevail, based on both the need for risk-reduction measures and realism about what hopefully can be negotiated in the foreseeable future.

In the nuclear arena, these negotiations can no longer be limited to the United States and Russia. China will have to be involved and perhaps other nuclear weapons countries as well—if not at the start, then later. Beijing continues to oppose nuclear-arms reductions, but it has an interest in preventing accidental nuclear war or a limited nuclear attack that escalates into an all-out exchange.

The new approach could use a new label to mark the changed NATO policy. The title could be Risk Prevention and Reduction (RPR) to underline the nature of the approach. To further explore concepts, measures, and a negotiating strategy, NATO could create an RPR advisory group, consisting of civilian and military experts from several allied countries. Within one year, this group could produce a report with concrete recommendations on issues like concepts and proposals. In the meantime, NATO arms control expert groups should sketch out new approaches, including by using the expertise of academia and think tanks.


1 “Harmel Report,” NATO, updated November 16, 2017,

2 “Reducing the Risks of Conventional Deterrence in Europe: Arms Control in the NATO-Russia Contact Zones,”OSCE Network of Think Tanks and Academic Institutions, December 2018,

3 Hans-Joachim Schmidt, “A Fresh Start of Conventional Arms Control in Europe Will Face Many Structural Problems,”Peace Research Institute Frankfurt, report no. 151, 2017,

4 “From the Baltic to Alaska: More Russian Air Provocations Reported,” Russia Monitor, Warsaw Institute, January 30, 2019,

5 “China Won’t Join Talks on Trilateral Nuclear Disarmament Deal: FM,” Xinhuanet, May 6, 2019,

6 Hans M. Kristensen and Matt Korda, “Status of World Nuclear Forces,” Federation of American Scientists, accessed May 16, 2019,

7 “Reducing the Risks of Conventional Deterrence in Europe,” OSCE Network of Think Tanks and Academic Institutions.