In a 2017 speech, Russian President Vladimir Putin stated that “artificial intelligence […] is the future of all mankind. […] The one who becomes the leader in this sphere will be the lord of the world”.[1] But despite the impact the ongoing artificial intelligence (AI) arms race is having on national military policies and international diplomacy, it enjoys a surprisingly low political profile.
A third revolution in warfare: the inconspicuous rise of autonomy
Because technological advances in autonomy are generally incremental, the fact that lethal autonomous weapons systems (LAWS) constitute a military revolution tantamount to that of the nuclear age seems to avoid the public spotlight. Credible risks associated with LAWS include the enhancement of terrorist tactics, empowerment of authoritarian rulers and vulnerability to bias, hacking and malfunction,[2] and states’ concerns with these weapons vary from dystopian outlooks to the prospect of more humane warfare.
A major challenge in the field of LAWS is the lack of consensus on how to even define “autonomous weapons”. The task of regulating weapons systems that – depending on the operational definition – may or may not yet exist, has proven difficult; autonomy is already being implemented in different parts of existing weapons systems.[3] One could argue that because the world has coped well without specific legal regulation of nuclear weapons, the same can be true for LAWS. However, such an attitude would greatly disregard the specific characteristics of LAWS that separate them from previous military revolutions.
The public paradox: lowering the threshold of war
Immanuel Kant once wrote that democratic peace “relies upon the public not supporting unnecessary wars as they will be the ones called upon to fight in them”.[4] This draws to mind the massive American public resistance to the Vietnam War, and is given practical meaning with the general decline in public opposition to ongoing wars when many states transitioned from a conscription or a draft-based military to an all-volunteer force.[5] Although warfare has seen many developments since Kant uttered these words, they may gain renewed relevance with the autonomous revolution.
Increasing autonomous capabilities in weapons systems reduces the need for soldiers both directly and indirectly participating in hostilities. This makes war cheaper, both in terms of human and economic resources. Such a development foreshadows a disconcerting future, where war could become preferable to costly and tedious diplomacy, particularly with regards to out-of-area operations where the intervening party has a clear military advantage. As for the public opinion, war as a political act that does not directly involve the national population could alienate them further from their governments’ actions on foreign soil.
As such, the autonomous revolution creates an even greater disparity in how civilian populations inside and outside of conflicts are affected by warfare – and quite possibly, gives rise to a greater indifference amongst the latter. That would be an unfortunate development.
International inefficiency
Discussions on LAWS have officially taken place under the United Nations umbrella since 2014,[6] but the forum has yet to produce anything of substance. A significant divide has emerged between the states calling for a ban on LAWS and the world’s leading military powers, such as the United States, the Russian Federation, Israel, and South Korea. While allegedly eager to discuss autonomy and promote transparent national policies on the matter, most of these countries oppose all forms of commitment. On the other side of the aisle is the Non-Aligned Movement (NAM), outnumbering these military powers with more than 100 states and consistent in their call for a ban. Yet the mighty few continue to be the constraint on – and consequently a necessary force behind – change.
States strive to preserve flexibility in their policy options, and are generally unlikely to make unilateral commitments. The NAM countries’ clear stance is unlikely to change this. A wild card in the pro-ban group is China, who in 2018 had a surprising change of heart and abandoned the anti-ban camp, despite their heavy investment into AI research.[7]The Chinese are however pursuing a ban solely on the use of LAWS – not production or sales – which illustrates the significant economic interests at play. A rapidly expanding global AI market makes the LAWS industry attractive for both state-run and private business, which entails that LAWS are going through a “widely unaccounted for development”.[8] Current projections show global military spending on AWS and AI to reach $16 and $18 billion respectively by 2025.[9]
The “nuclear argument” of political stability through Cold War-esque principles like balance of power and deterrence holds little credibility in today’s conflict image; while states argue definitions, development advances and LAWS become available to third-party actors. Although AWS technology is per now concentrated in powerful and wealthy states, Moore’s law (price drop over time), declining costs of production and 3D printing will soon enable non-state actors to procure similar technologies.[10] Such tendencies are already visible; in 2019, quadcopter drones cost as little as $25, and a small $35 Raspberry Pi computer could run AI advanced enough to defeat American Air Force fighter pilots in combat simulations.[11] Houthi rebels are already using weaponized drones, and both ISIS and Boko Haram have adapted drones for use as improvised explosive devices. Simply put, Western states are ensuring the startup costs for these actors.
The grim reality is that widespread agreement between hundreds of minor states means nothing if the major military powers remain unwilling. And while states argue the nitty-gritty of terminology well below the public radar, technology advances and spreads to irregular actors and rogue states. Whichever opinion one might have on the current developments, they should be subject to public debate. If not, we might suddenly find ourselves in the midst of an irreversible development where human control in warfare is gradually lost.
This article is written by Mathea Reine-Nilsen, deputy leader of YATA Oslo. She holds a bachelor’s degree in international relations from the University of Oslo and is currently enrolled in law studies at the same university. The views expressed in this article are entirely the views of the author, and do not necessarily represent the views of YATA Norway.
[1] Slijper, Kayser and Beck for Pax, “State of AI. Artificial intelligence, the military and increasingly autonomous weapons”, 2019, 4. Available at https://www.scribd.com/document/409149044/State-of-Artificial-Intelligence-Pax-Report.
[2] Garcia and Haner, “The Artificial Intelligence Arms Race: Trends and World Leaders in Autonomous Weapons Development”. Global Policy Volume 10, Issue 3 (2019), 331.
[3] Ibid.
[4] Ibid, 332.
[5] Ibid.
[6] Reports and other documents from the conferences can be found at https://www.unog.ch/80256EE600585943/(httpPages)/5535B644C2AE8F28C1258433002BBF14?OpenDocument.
[7] Pax, “State of AI”, 13.
[8] Garcia and Haner, “The Artificial Intelligence Arms Race”, 331.
[9] Ibid.
[10] Ibid.
[11] Ibid, 331-332.