NUCLEAR DETERRENCE IN THE AGE OF AI

Awais Siddique

Emerging technologies and disruptive innovations have been a key influencer of international relations and security, and an integral defining characteristic of the major powers. Incorporation of new technologies in the realm of the military often result in new threats and challenges, which challenge the conventional notions of competition and power, changing the security landscape altogether. Among all emerging technologies, Artificial Intelligence (AI), although in its embryonic phase, has been most potent and torrential due to its crosscutting utility in military and intelligence affairs. The most obvious application of AI in military and intelligence affairs is its use in developing smart and interconnected systems and its role in decision-making. At one end, AI acts as the status-quo enhancer but on the other hand, it reverses the balance of power. The nascent developments in the field of AI is a major threat to the prevalent nuclear deterrence of nations because of the ability of AI to make swift analysis, its utility in payload delivery, and the most significant one is the use of AI in the robots –which could be used for civilian and military purposes.

Artificial Intelligence (AI) is “the capability of a machine to imitate intelligent human behaviour” and experts believe that the superintelligence of AI will surpass human intelligence in the next few years. Current developments in the field of AI is rapidly growing, having far-reaching consequences for national and international security and dominant military powers like the USA, China and Russia are harnessing this disruptive technology for their military modernization and defence purposes. The use of AI in intelligence gathering and analysis, logistics, AI-based weapon platforms, cyber and information related operations, command and control, and in a variety of semiautonomous and autonomous vehicles is a bone of contention in the academia, supranational organizations striving for peace and militaries struggling to safeguard their sovereignty and national interests.

The military application of AI includes “machine learning, neural networking, and autonomy that broadly feature in robots, transportation, communication, Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance (C4ISR) systems, and nuclear arsenals”. These AI-led advances in the domain of the military render the technology as a force-multiplier and heighten the probability of AI exploitation for threatening the adversarial nuclear forces and escalating tensions.

The alarmists contend that AI winter is unlikely and the superintelligence of AI has the potential to disturb the prevalent nuclear deterrence. The proliferation of AI-based anti-ballistic missile systems and drones are diminishing nuclear deterrence. Coercive use of AI can lead to strategic imbalance, challenge nuclear deterrence, and compel humans to make devastating decisions, intentionally or inadvertently. The current trajectory of AI development, together with that of complementary information technology and other advancements, indicate that in near future AI militarization will also largely affect the nuclear-security and deterrence. A report published by Rand Corporation in 2018 construes that, the military use of AI could intensify the risk of nuclear war and severely compromise the prolonged nuclear strategic stability. It also delineates that “AI has significant potential to upset the foundations of nuclear stability and undermine deterrence by the year 2040, especially in the multipolar strategic environment”. The militarization of AI, especially the proliferation of Lethal Autonomous Weapon Systems (LAWS) or subversion of an adversary’s AI capability, can also increase the possibility of an armed escalation between states leading to nuclear use during a crisis.

Conversely, the complacent view is that an AI competition can lead to strategic stability and bolster deterrence, as the machine has no bias and cannot make mistakes, thus through the use of AI for situational awareness in defence, there are fewer chances of human error and risks of miscalculation, which can foster stability. The propensity of military use of AI will less likely aggravate the prevalent nuclear stability and deterrence of multipolar nuclear world order. Nuclear-armed states leveraging AI to achieve or sustain first-mover advantages in this multipolar context will likely destabilize this fragile order with uncertain outcomes.

Having its pros and cons, at one end, the application of AI for military modernization allows the weaker nuclear-armed states to overcome the imbalance of power, while on the other side it also exacerbates suspicions that dominant military powers may employ it to further solidify their martial supremacy and engage in more provocative actions.

As Alfred Thayer Mahan hinted in the last century that, “whoever rules the waves rules the world”, Russian President Vladimir Putin succinctly puts it now as, “whoever wins the race for AI superiority will become the ruler of the world.” The Vision 2020 and 2025 of the world’s most powerful and advanced military, the US, acknowledged the incorporation of AI for offence and defence purposes, which highlights the significance of technology in military affairs.

Carl Von Clausewitz describes war as “having an enduring nature but a constantly changing character”. In this lieu, it is high time that all nation-states, especially nuclear-armed ones, should consider the latent, disruptive and inevitable impact of AI on nuclear security before it becomes extremely challenging to preserve strategic stability. In the coming decades, AI will more likely erode the strategic significance of nuclear deterrence and nuclear powers will need institutional reassurance to help limit the nuclear risk. Nuclear restraints, in the coming quarter-century, will necessitate the cooperation of rival states underlying a fortuitous combination of technological, military, and diplomatic measures.

Meanwhile, in the South Asian region, India and Pakistan are progressively concerned about the potential benefits of AI for defence and security purposes. Currently, a debate has started on the opportunities and risks posed by the AI renaissance in the military realm. However, it suffers from large loopholes, particularly in the incipient discussion on the potential impact of AI on strategic stability and nuclear risk in South Asia. A glance at the military modernization plans of India and Pakistan hints that both South Asian nuclear powers are playing catch-up in the world competition on military AI. A comparison of Indian and Pakistani military advances in the realm of military AI with the dominant powers like the US, China, and Russia shows uncertainty and both nuclear archrivals seem to be under-resourced and inefficient in defence research and advance tech industries.

It is evident from the current developments that AI will be an imperative part of military affairs in near future and will have a disruptive impact on the security landscape. Although current military affairs are more automatic rather than autonomous and AI application is limited, such as semi-autonomous or teleoperated robots and UAVs but there is an ongoing global effort to develop such systems for defence purposes such as autonomous tanks, underwater vehicles etc. which in turn is spiralling into an arms race of autonomous systems. It is the need of the hour to put halt to the arms race of autonomy and incorporation of disruptive technologies like AI in military affairs and this is possible through taking confidence-building measures and increased engagement of nation-states in multilateral forums and global institutions like the UN Group of Experts on Lethal Autonomous Weapon Systems.

Keeping nuclear peace in a time of such technological advances will require the cooperation of every nuclear power in the domain of emerging technologies especially AI and its use as a weapon. It will require new global institutions and agreements, new understandings among rival states, and new technological, diplomatic, and military safeguards. For this purpose, COPAIR also organized a webinar for conceptual clarity and understanding the implications of AI on the nuclear deterrence of Pakistan. This webinar was part of the series of conferences COPAIR will organize in future and they will not only highlight the threats of AI to national security and nuclear deterrence but will also delineate the requisites for the incorporation of AI in the broader military modernization agenda. The resolve behind these activities are to provide recommendations to the policymakers and address the challenges of national security and modern warfare spurring from the military use of AI.

Newsletter SignUp

Awais Siddique
About Awais Siddique 9 Articles
Assistant Editor TAT and Digital Editor at Melange International Magazine