New Technologies and Global Security from the Perspective of Global Security Governance

BY DR.JOHN BORRIE

Researcher with the United Nations Institute for Disarmament Research (UNIDIR), Head of WMD and other Strategic Weapons Programme/Research Coordinator in Geneva

The technologies discussed are anti-ballistic missile defenses, hypersonic and other advanced long-range weapons, anti-satellite weapons, un-crewed weapon systems, cyber,artificial intelligence and machine learning.Machine learning, although not a new technologyper se— yet, is a more “usable”weapon.Individually or in combination with each of these developments has implications for the nuclear balance or will have actual effects in crisis situations that may be hard to predict, or both.

Taken together, these technologies are one of four sources of growing strategic unpredictability-making the world a more dangerous place.The other sources are more nuclear-armed states uncertanties; tenser relations between some of these states; and,lastly, the deterioration of the fabric of international institutions and norms that contribute to stability.In this regard, the weakening of the “arms control enterprise” can be seen in stalled bilateral nuclear reductions between the United States and Russia.2At the multilateral level, the Conference on Disarmament (CD)is still deadlocked after more than two decades and the Nuclear Non-Proliferation Treaty (NPT)regime is struggling to contain political divisions exacerbated by lack of progress on nuclear disarmament.

As bleak as the current outlook is, arms control approaches on new strategic technologies could contribute to stability as in the past, greater strategic dialogue,clarifications of nuclear use doctrines and crisis communication improvements could help strengthen nuclear weapon risk reduction measures, and in turn could contribute to the trust necessary for states to agree to resume progress toward a nuclear-weapon-free world.

Several main technological developments are relevant to thinking about two problems in global security.The first issue concerns their implications for the nuclear balance, which makes these technologies both products of—and fodder for—strategic modernization and arms racing among major strategic competitors like the United States, Russia,China, and India.The second is that although several of these technologies are ostensibly intended to strengthen deterrence, in crisis escalation situations their actual effects may be difficult to predict and contribute to greater crisis and instability instead.These are important reasons why intensified efforts toward strategic arms control agreements are needed in the two major triadic nuclear relationships: U.S.-Russia-China, and China-India-Pakistan.In addition, without strengthening the multilateral disarmament,arms control and non-proliferation in these strategic relationships, it is difficult to prevent powers conflict and nuclear war.

I.New technology and nuclear instability

After the United States detonated atomic bombs in Japan in 1945, nuclear weapons are recognized as something qualitatively different from all previous strategic technologies.No means existed to effectively counter them, and nuclear strategy proceeded from this dreadful and inescapable fact.Although “Mutually Assured Destruction” was never official U.S.policy, yet it captured the sentiment of nuclear deterrence due to the absolute certainty of catastrophically damaging nuclear reprisal.The 1972 Anti-Ballistic Missile (ABM)Treaty between the United States and the Soviet Union in effect acknowledged this.While it is held,the ABM Treaty removed the risk that one side would lose its retaliatory nuclear capability to the other due to missile defense developments.It thus served to open the way for strategic nuclear reductions.

Today, nuclear war planners are less certain than they were about that “strategic stability”.The possibility of new technology leading to nuclear vulnerability that forms the basis for nuclear deterrence is a vexing issue for them.Nuclear planning is based on worst-case contingencies.Certain contingencies that would have major ramifications cannot be excluded(say, a “bolt-from-the-blue” or “decapitated”attack in which national nuclear command-and-control systems were to be destroyed, nullifying one’s nuclear deterrent),then steps have to be taken to forestall them.This is one reason why so much nuclear“modernization” has an element of qualitative improvement to it.It also helps to explain why arms racing dynamics can take hold among states.And it also means some strategic technologies developed to enhance the credibility of one’s nuclear deterrent capability can, in crisis situations, potentially be destabilizing.

II.Which technologies are of strategic concern?

a.Anti-ballistic missile defences

Since it left the ABM Treaty in 2002, the United States has developed more advanced missile defense systems to intercept missiles on ballistic flight paths.Progression developing reliable homeland missile defense has so far remained limited.However, missile defense at the tactical and theatre levels focused on missions to protect military assets and troops have become capabilities of importance to more countries as the technology has improved.The United States, Russia, India, France, Israel and China have all developed missile defense systems, and the United States has made systems available to its allies including Saudi Arabia in the Middle East, and Japan and South Korea in Asia.

As ballistic missile defense systems advance they are becoming more integrated‘systems-of-systems’.It means previous distinctions between systems intended to intercept non-nuclear missiles and nuclear-armed missiles may be blurring.Russians say it is concerned that missile defense capabilities deployed in Eastern Europe could undermine its second strike nuclear retaliatory capability.In March 2018, Russian President Vladimir Putin cited a new generation of Russian strategic nuclear systems as stemming from U.S.withdrawal from the ABM Treaty.3China may conceivably share this concern in Asia due to its relatively small number of deployed nuclear missiles.

b.Hypersonic and other advanced,longer-range weapons 4

Hypersonic weapons travel through the atmosphere at more than five times the local speed of sound.Several states are actively pursuing novel long-range manoeuvrable weapons, most significantly hypersonic boost-glide systems equipped with hypersonic glide vehicles (HGVs).HGVs are unpowered after separation and do not follow a ballistic flight path after the boost phase.They may have an enhanced ability to overcome missile defenses due to both this manoeuvrability as well as their depressed trajectories relative to standard ballistic missiles.Today, the United States, Russia, China, and most recently France have active HGV acquisition programs.Other states are reportedly interested in the technology.The U.S.intends to use boost-glide technology with conventional or kinetic(non-explosive)warheads; it is unclear whether Russia and China’s systems will be nuclear-armed.

Although the military utility of HGVs remains unclear, an arms race dynamic is unfolding in pursuit of this technology.Possible ambiguity about the nature of an HGV’s warhead (nuclear or conventional), together with uncertainty about its intended target,means the potential for strategic misunderstanding is considerable.Missiles launched carrying HGVs could be interpreted as signaling an imminent nuclear attack.Even if HGVs are subsequently shown on impact to be conventional it could also prompt “use it or lose it” dilemmas for nuclear-armed states at risk of being targeted if they believe these weapons have been directed against their nuclear early warning, command and control infrastructure.In view of these ambiguities it is conceivable that the advent of HGVs will prompt some nuclear-armed states to amend their doctrines to expand the conditions necessary for the use of nuclear weapons in response to HGV deployments, or/and by placing their nuclear forces on higher alert.

Other developments in advanced,longer-range missiles may also put pressure on nuclear doctrines and war plans.Russia and China have long asserted that conventional cruise missiles could upset the strategic balance.5In recent years, the United States has produced and exported to some allies the air-to-ground missiles, a low-observable air-launched conventional cruise missile it first used in Syria in April 2018.A new version of this missile with an extended range of around 1,000km is entering service.6Launched from bombers outside an adversary’s most heavily-defended airspace, such cruise missiles could evade air defences due to their stealthy characteristics to strike strategic targets such as nuclear command and control systems and mobile nuclear launchers.A related issue, as in HGVs, is that strikes with conventional cruise missile systems may be misidentified as air-launched nuclear-tipped cruise missiles by the targeted state.7

c.Anti-satellite weapons 8

States have long appreciated the importance of outer space for many military functions ranging from communications to surveillance to early warning of nuclear attack, and many military operations rely on satellite access today.The 1967 Outer Space Treaty prohibits the stationing of nuclear weapons in space,although a proposal by China and Russia for a treaty to prevent placement of other weapons there has not achieved widespread attraction.9Western countries claim it would not be verifiable, and that in any case it is not necessary to put weapons in space to pose a danger to other space objects.Indeed, three countries to date (the United States, China and,in April 2019, India)have tested ground-launched anti-satellite (ASAT)interceptor capabilities.10However, it is not only such missile interceptors, but also a variety of “non-kinetic” cyber and electronic counter-space capabilities that can disrupt or destroy satellites.11And any space object is at risk from collision.This makes the increasing ubiquity of co-orbital drone technology the significant international concern,especially as it is difficult to identify the intentions of proximity manoeuvres until collision is imminent.

Why are anti-satellite weapons of strategic concern? The most obvious reason is that various countries are increasingly reliant on space high-tech for civil and military purposes.Even if states have not placed objects in space that are unambiguously weapons and have not deliberately collided with or blown up each other assets yet, non-kinetic offensive space operations have been going on for some time.12The formation of national “space forces” of the United States and India reflects concerns as to their vulnerability of space-based infrastructure,but may also contribute to more overt competition and arms racing dynamics.If space-based systems to detect missile launches and flight paths are deployed, it will in turn increase the desire for ASAT capabilities.However, debris created by the destruction of space objects using direct-ascent ASATs would likely cause major disruption and could render space effectively unusable for some time, with major consequences for daily life on earth for billions of people.13

d.Un-crewed weapon systems

In 2013, the United Nations Human Rights Council’s Special Rapporteur presented two important reports.One concerned armed unmanned aerial vehicles (UAVs, or armed drones),14and the other focused on so-called“lethal autonomous robotics”15(LARs).These brief papers were influential in prompting international concern on these topics.First, there is international disquiet over the role of remotely-piloted UAVs in clandestine operations outside conventional battlefields in which it is not always clear which legal rules apply.Yet these concerns have not yet crystalized into focused discussions in an international forum.16In contrast, these topics were mentioned in discussions of the UN Convention on Certain Conventional Weapons,and these discussions continue.Questions raised there include what the implications are of machines selecting targets and attacking human beings.Whether such systems can be sufficiently predictable or reliable, and whether autonomous weapons are ethical.17

Although considered separately to date,both armed drones and autonomous weapon systems have yet produced significant impact,but could create similar issues in situations of strategic escalation or crisis.From an operator’s perspective both kinds could be more attractive for some military roles as they are more expendable than crewed platforms(for instance, able to be sent into contested airspace or waters to gather intelligence even if likely to be intercepted and destroyed).However, there is evidence to suggest persistent differences between how users and those on the receiving end would view the use of un-crewed systems in this manner.18These differences could result in misunderstanding and further escalation.19After one incident involving a Chinese drone in 2013, Japan stated it had new rules of engagement to shoot down any further drones incurring on its airspace.China stated that any attack on its drones would be an “act of war” and that China would “strike back” if that happened.20And despite the best efforts of designers and operators, un-crewed aerial systems have higher rates of acting unpredictably and of crashing.These can add additional variables in tense situations or crises.

e.Cyber

Modern life, including modern military systems, depends on digital data created, kept,managed and moved around on cyber networks.The exploitation of code for hacking, spoofing,phishing, stealing, disrupting and even altering or destroying data has moved in from the margins to become a major security focus over the last decade.Now, despite the difficulty of attributing hacks, it is a poorly kept secret that several of the major military powers undertake offensive cyber operations against each other,as well as others.21Cyber offensive operations are increasingly ubiquitous and persistent.The lines are blurry between state versus-state offensive operations, espionage and other activities, including theft and extortion, in which civilian infrastructure and bystanders are victims.

Cyber offensive capabilities rise to the level of strategic concern in two kinds of scenario.The first is hacking or other cyber interference with nuclear early warning, command and control or decision support systems—or creating fear in a target country.The target country claims that it is under cyber threat,which could lead to nuclear escalation;in extremis, nuclear “use it or lose it” scenarios are even conceivable.The second kind of scenario concerns those in which an aggressor uses cyber offensive means to disable the critical societal infrastructure an adversary’s population relies on.22In June 2019 it was reported that both the United States and Russia were penetrating deeply into each other’s electric utilities, planting malware potentially capable of disrupting their national power grids.23It is significant then, that in its 2018 Nuclear Posture Review, the United States pointedly refused to rule out a nuclear response to “non-nuclear strategic attacks” like a major cyber-attack.24Thus, cyber threats already appear to be impacting the nuclear use doctrines of various states.

f.AI and machine learning 25

Dramatic advances in artificial intelligence(AI)are having wide-ranging societal impacts.Algorithm-based machine systems are becoming vastly better at self-optimizing their performance based on various techniques,many of them related to pattern recognition and matching of data.This will improve the ability of machine systems to perform various critical military functions with a greater level of autonomy ranging from communications and logistics, to network defense, fire control,intelligence analysis support and even the selection of targets and launching of attacks.Concerns are being expressed about the emergence of an ‘AI arms race’ between the United States and China or an “AI cold war that threatens all” in particular.26

Concerns that AI will exclude humans from decision making in conflict are probably overblown, at least for now.As one recent study found, there are greater risks from machine learning-powered AI applications and autonomous systems being too brittle in their design: “They may fail spectacularly when confronted with tasks or environments that differ slightly to those they were trained for.Their behavior is also unpredictable as they use algorithms that are opaque.It is difficult for humans to explain how they work and whether they include bias that could lead to problematic—if not dangerous—behaviors.They could also be defeated by an intelligent adversary through a cyber-attack or even a simple sensor spoofing trick.An immature adoption of the latest developments of AI in the context of nuclear weapons systems could have dramatic consequences.”27

At the same time, systems of various kinds based on algorithms are increasingly indispensable to militaries.For two reasons in particular, needs to be strengthened over increasing reliance of military command chains on automation for early warning and detection of nuclear or other strategic attack.First, with such techniques advance in the nuclear context,prospective capabilities in machine learning and other AI-related techniques could affect assured nuclear retaliatory capability and thus current strategic balances.28Second, the uptake of AI will be a continuous process with heterogeneous impacts on human decision-making processes in nuclear early warning, command and control systems, even if the final decision to launch nuclear weapons always lies with humans.Yet a lesson from crises in the Cold War is that such human contextual awareness was crucial in averting nuclear use on several occasions.29Will human decision makers be able to make the right call when they depend on AI?

g.More “usable” nuclear weapons

While not strictly speaking a new technology, more usable nuclear weapons are mentioned here because they are becoming part of the mix of strategic technologies,30to this end, there are discussions within some nuclear-armed states about deploying “more usable” nuclear weapons.It seems intended that these weapons could be tactically useful and might even scare an adversary.31

Deploying nuclear weapons in these ways would in fact most certainly be very destabilizing.At present, nuclear-armed states try to be very careful in terms of the roles they accord to nuclear weapons—emphasizing deterrence and nuclear weapon use only in extreme cases of existential self-defense.This has contributed to a nuclear non-use norm.More usable nuclear weapons promise to undermine the non-proliferation norms like those embodied in the NPT.In peacetime it requires changes in doctrines of nuclear use,and such changes to doctrine could add ambiguity to crisis situations, and in effect lower the threshold to nuclear use.

III.Why new technology demands arms control

Each of the developments outlined above has implications for the nuclear balance or will have actual effects in crisis that may be very difficult to predict, or both.In all cases this has implications for global security, since there are few things more “global” than a nuclear crisis or use of nuclear weapons.

Of course, it is hard for scholars or decision-makers to foresee the precise ways in which these technologies and other factors will interact.Of course, it is impossible to know whether any will offer any of the world’s major strategic rivals a significant or lasting comparative advantage over others.In the absence of strategic trust, this uncertainty tends to contribute to arms racing and further instability.How can the world’s major strategic competitors find a way out of this mess?

One fundamental process at work is the increasing speed of modern warfare, driven by the computer microprocessor and the relevant technologies, and which often serve to compress human decision-making time.There is a pressing need to ensure that, where the use of nuclear weapons is concerned, human decision-makers are not simply swept along with the tide in a crisis.Dialogue is needed between policy makers from different governments developing—or thinking of developing—the technologies discussed in this paper to ensure that they understand the implications and can account for differences in how others see these.

This kind of exchange could be part of strategic dialogue oriented toward nuclear weapon risk reduction pathways and measures.Risk reduction frameworks, elements and pathways are already something that research institutes like UNIDIR focus on,32and nuclear risk reduction has recently attracted broad support in discussions contexts like the NPT.As part of this, attention should be paid to the implications of new technologies discussed here for nuclear doctrine, strategy, operations,and transparency that goes much further than the periodic exchanges seen among the five NPT nuclear-weapon states although these are welcome.33There are a range of ideas for practical measures that could help reduce nuclear risk, including better means of crisis communication not only bilaterally, but in the context of the U.S.-Russia-China and China-India-Pakistan strategic triads.

Nuclear risk reduction measure can help to pave the way toward the development of specific arms control measures, which could capture aspects of the new technologies discussed in this paper and contribute to greater strategic trust.The goal must be to resume progress toward a nuclear-weapon-free world,which the United Nations Secretary-General recently called for in his May 2018Agenda for Disarmament.34While this might seem utopian to some, it is more utopian to believe that nuclear deterrence can operate indefinitely without an accident, miscalculation or other situation leading to nuclear use, especially in today’s more complex and changing strategic landscape.And it is sometimes forgotten that arms control was born in the 1950s during a nuclear arms race in which new technologies—the hydrogen bomb,inter-continental ballistic missiles,satellites—were being introduced and trust between the two superpowers of the time was low.Arms control has always been a matter of enlightened self-interest, whether for the purposes of “stability” or more durable security conditions: today, as the strategic balance has shifted it also requires initiative and leadership beyond Washington and Moscow.China has an important, constructive role to play.

Footnotes:

1.Dr.John Borrie is Head of Weapons of Mass Destruction and Other Strategic Weapons Programme.The views expressed here are the author’s own, and do not necessarily reflect those of UNIDIR, its research funders,or the United Nations.The author thanks Dr.Pavel Podvig, Dr, Ryan Snyder and Daniel Porras and for their critical feedback on a draft of this paper.

2.Lewis A.Dunn,Reversing the Slide: Intensified Great Power Competition and the Breakdown of the Arms ControlEndeavour, Geneva, UNIDIR, February 2019:.

3.Vladimir Putin, ‘Presidential Address to the Federal Assembly’, 1 March 2018:.

4.Some of this section is drawn from John Borrie, Amy Dowler and Pavel Podvig,Hypersonic Weapons: A Challenge and Opportunity for Strategic Arms Control(A Study Prepared on the Recommendation of the Secretary-General’s Advisory Board for Disarmament Matters), New York, United Nations, February 2019:http://unidir.org/files/publications/pdfs/hypersonic-weapons-a-challenge-and-opportunity-for-strategic-arms-contr ol-en-744.pdf

5.Nikolai N.Sokov, “Modernization of nuclear weapons: How it influences strategic stability”, Moscow, PIR Center, 13 December 2018: .

6.Oriana Pawlyk, “U.S.May Ramp Up Buy of the Missile That Just Made Combat Debut in Syria”,Defense News,17 April 2018:

7.As of 2017, the U.S., France and Russia reported stocks of nuclear-armed cruise missiles.See Christine Parthemore, “The unique risks of nuclear-armed cruise missiles”, Chapter 4 in John Borrie, Tim Caughley and Wilfred Wan (eds.),Understanding Nuclear Weapon Risks, Geneva, UNIDIR, 2017, p.45.See also Sokov, op cit(2018).

8.Daniel Porras,Briefing paper for the United Nations Disarmament Commission–Shared risks: An examinationofuniversalspacesecuritychallenges, Geneva, UNIDIR, February 2019:.

9.Treaty on the Prevention of the Placement of Weapons in Outer Space, the Threat or Use of Force against Outer Space Objects (PPWT).

10.Russia has not tested a direct-ascent ASAT but is generally believed to have at least a limited capability.See Brian Weeden and Victoria Samson (eds.),Global Counterspace Capabilities: An Open Source Assessment,Washington D.C., Secure World Foundation, April 2019, p.2-21.

11.Rajeswari Pillai Rajagopalan,Electronic and Cyber Warfare in Outer Space(Space Dossier 3), Geneva,UNDIR, 2019: .

12.See Weeden and Samson, op cit(2019).

13.Daniel Porras,Towards ASAT Test Guidelines (Space Dossier 2), Geneva, UNIDIR, 2018:http://unidir.org/files/publications/pdfs/-en-703.pdf

14.Christof Heyns,Report of the Special Rapporteur on extrajudicial, summary or arbitrary executions, United Nations Document A/68/382, 13 September 2013: .

15.Christof Heyns,Report of the Special Rapporteur on extrajudicial, summary or arbitrary executions, United Nations DocumentA/HRC/23/47, 9 April 2013:.

16.John Borrie, Elena Finckh and Kerstin Vignard,Increasing Transparency, Oversight and Accountability of ArmedUnmannedAerialVehicles, Geneva, UNIDIR, October 2017:.

17.See UNIDIR’s series of briefing papers on these and other questions related to the autonomization of weapon systems:.

18.J.Schaus and K.Johnson,Unmanned Aerial Systems’ Influences on Conflict Escalation Dynamics(CSIS Briefs), 2 August 2018,.

19.George Woodhams and John Borrie,Armed UAVs in conflict escalation and inter-State crisis, Geneva,UNIDIR, 2018,pp.7-8:.

20.Paul Scharre,Army of None: Autonomous Weapons and the Future of War, New York & London, W.W.Norton,2018, p.208.

21.China, Iran, Israel, North Korea, Russia, the UK and U.S.all have active cyber offensive operations capabilities.See David Sanger,The Perfect Weapon:War, Sabotage and Fear in the Cyber Age,Melbourne &London, Scribe, 2018.

22.Gordon Correra, “NHS cyber-attack was 'launched from North Korea'”,BBC World News, 16 June 2017:.

23.David E.Sanger and Nicole Pertroth, “U.S.Escalates Online Attacks on Russia’s Power Grid”,New York Times, 15 June 2019: https://www.nytimes.com/2019/06/15/us/politics/trump-cyber-russia-grid.html.

24.“The United States would only consider the employment of nuclear weapons in extreme circumstances to defend the vital interests of the United States, its allies, and partners.Extreme circumstances could include significantnon-nuclear strategic attacks.Significantnon-nuclear strategic attacks include, but are not limited to,attacks on the U.S., allied, or partner civilian population or infrastructure, and attacks on U.S.or allied nuclear forces, their command and control, or warning and attack assessment capabilities.” See U.S.Department of Defense,2018 Nuclear Posture Review, p.21:.

25.This section is based on John Borrie, “Cold war lessons for automation in nuclear command and control systems” in Vincent Boulanin (ed.),The Impact of Artificial Intelligence on Strategic Stability and Nuclear Risk VolumeIEuro-AtlanticPerspectives, Stockholm, SIPRI, May 2019,pp.41-52:.

26.Remco Zwetsloot, Helen Toner and Jeffrey Ding, “Beyond the AI arms race: America, China, and the dangers of zero-sum thinking”,Foreign Affairs, 16 November 2018; and Nick Thompson and Ian Bremmer, “The AI cold war that threatens us all”,Wired, 23 October 2018.

27.Boulanin,op cit, p.xi.

28.Geist and Lohn (note 3).

29.Patricia Lewis et al,Too Close for Comfort: Cases of Near Nuclear Use and Options for Policy, London,Chatham House, 2014:.

30.Julien Borger, “Nuclear weapons: experts alarmed by new Pentagon 'war-fighting' doctrine”,The Guardian, 19 June 2019: .

31.Nikolai N.Sokov, “Why Russia calls a limited nuclear strike “de-escalation”,Bulletin of the Atomic Scientists,13 March 2014: .

32.See Wilfred Wan,Nuclear Risk Reduction: The State of Ideas, Geneva, UNIDIR, April 2019:; and John Borrie,Tim Caughley and Wilfred Wan (eds.),Understanding Nuclear Weapon Risks, Geneva, UNIDIR, 2019:.

33.China, France, Russia, the United Kingdom and United States.

34.https://www.un.org/disarmament/sg-agenda/en/.