Abstract

The legality of autonomous weapon systems (AWS) under international law is a swiftly growing issue of importance as technology advances and machines acquire the capacity to operate without human control. This paper argues that the existing laws are ineffective and that a different set of laws are needed. This paper examines several issues that are critical for the development and use of AWS in warfare. It argues that a preemptive ban on AWS is irrelevant at this point and urges the appropriate authorities to develop a modern legal framework that is tailored to embrace these state-of-the-art weapons as the Law of Armed Conflict (LOAC) develops. First, this paper explores the myriad of laws designed to govern the potential future development and deployment of artificial intelligence and AWS in the context of International Humanitarian Law or LAOC. Second, the paper argues that it will be challenging for AWS to fulfill the requirements laid out under the International Committee of the Red Cross and LOAC for the rules of humanity, military necessity, distinction, proportionality and precaution, especially as it is related to noncombatants. Third, the paper discusses command responsibility and argues that states should establish accountability for wrongful acts committed by the AWS. Finally, this paper contends that there is an urgent need for a new legal framework to regulate these AWS and presents different solutions for the legal framework of AWS.

Introduction

The legality under international law of autonomous weapon systems (AWS) is an issue that is growing swiftly in importance as technology advances and machines acquire the capacity to operate without human control.¹ This paper addresses several issues that are critical for the development and use of AWS in warfare and urges appropriate authorities to develop a modern legal framework that is tailored to embrace these state-of-the-art weapons as LOAC develops. Throughout human existence, humans have developed and continually progressed their use of tools and technology. Transhumanism, or the human desire to acquire new capacities, is as ancient as our species itself.² Humans have always sought to expand the boundaries of their existence, be it socially, geographically, or mentally.³ Human beings share a common tendency to search for a way around every obstacle and limitation to life and happiness.⁴ Humans want their gadgets to cook, clean, read, dictate, count, and solve problems for them.⁵ Now, humans must decide if they want gadgets to fight for them as well.⁶

Imagine a world where soldiers no longer return home in flag-draped caskets to heartbroken families, parents no longer grieve the loss of their children, and children no longer grieve the loss of their parents.⁷ This is a world where warfare is fought with autonomous robotics.⁸ Autonomous robots – mobile machines that can make decisions, such as to fire upon a target, without human intervention – can replace the human soldier.⁹ Robots can take the place of soldiers in dangerous missions, like tunneling through dark caves in search for terrorists, securing urban streets rife with sniper fire, patrolling the skies and waterways where there is little cover from attacks, clearing roads and seas of improvised explosive devices (IEDs), surveying damage from biochemical weapons, guarding borders and buildings, controlling potentially-hostile crowds, and even as frontline infantry.¹⁰ Weapon systems are becoming increasingly automated, and arguably some autonomous military systems have been deployed for years.¹¹ Recent advances in automated systems and the possibilities they portend have generated interest and anxiety within some militaries and defense ministries, and have also generated a movement of non-governmental activists seeking to ban fully autonomous weapons.¹²

This struggle between the interest in technological advancement and anxiety is not new to the world of warfare. The human desire for better and more powerful weapons has been driving human beings since the beginning of time. It may have started with replacement of fists with clubs, arrows with swords, and eventually bullets with cannons and other heavy artillery, but the outcome has not eliminated wars. Technological advancements created enthusiasm for war simultaneously with fear and dread of the destruction war will bring. The resulting carnage and killing is far beyond what any club bearing ancestor could have imagined. These technological advancements included tanks and air attacks, which kill hundreds or thousands in one strike. From these advancements, cities were destroyed and masses were killed. Technology advancements brought about the creation of nuclear weapons, which can reduce an entire city and wipe out its population with only one bomb. Whole cities have disappeared, affecting present and future generations. All these modern methods of fighting wars, at all levels of conflict, have required active human participation and involvement. In the wake of the September 11, 2001 terrorist attacks, there has been a dramatic upsurge in the presence and use of unmanned aerial vehicles, commonly identified by the media and the public as “drones.”¹³ This marks the beginning of the shift in human participation, as drones began replacing humans in the battlefield; however, they still require humans to operate them. The drone's rapid combat application not only gave rise to various legal issues and implications, but also set the stage for the development and eventual use of fully AWS.¹⁴ Professor Stuart Russell from the University of California at Berkeley said that these deadly drones could be the final stage of the ongoing technological march toward the AWS.¹⁵ Some scientists say arm makers have crossed into troubling territory; they are developing weapons that solely rely on artificial intelligence without benefit of human intuition and experience.¹⁶ This concern is growing as Artificial Intelligence (A.I.) installed in robotic drones will soon allow drones to perform targeted killing without the consultation of their human masters, working autonomously, coldly responding to a set of predetermined criteria.¹⁷

Unfortunately, International Humanitarian Law (IHL) or Law of Armed Conflict (LOAC) does not currently have provisions for this kind of technology, and it remains to be seen whether the international community would be supportive of a treaty that would limit or ban laws.¹⁸ However, the Convention on Conventional Weapons, which is grounded in IHL, provides an important framework to further understanding of the technical, legal, ethical and policy questions raised by the development and use of AWS in armed conflicts.¹⁹ Clearly, it is not too late to do this as the systems are already in widespread use by our military. Currently, technology is outpacing the extent of our laws. Our current system permits the creation of and use of technology where no laws have been enacted to address the concerns created by the new technologies. Thus, the current system is to create laws and regulations after the fact, when the technology is already in use in the public sphere. Militarized drones are no exception to this system. For example, drones were developed first, and then the laws and regulations were discussed. Accordingly, drones affected privacy laws, surveillance laws, and international laws regarding the killing of suspected terrorists and the LOAC in actual warfare. Rather than continue a reactive approach, societies need to be proactive and use careful thought out direction to legislate the laws and regulations required for the AWS before employment by nation states.

This paper argues that the existing laws are ineffective and a different set of laws are needed. In addition, this paper addresses several issues that are critical for the development and use of AWS in warfare. First, this paper explores the myriad of laws designed to govern the potential future development and deployment of A.I. and AWS in the context of LOAC. Second, this paper examines the challenges for AWS to fulfill the requirements laid out under the ICRC and LOAC for the preservation of human life, especially as it relates to noncombat. The paper argues that states should establish accountability for wrongful acts committed by the AWS. Fourth, the paper contends that all the nations involved in the development of A.I. and AWS must be party to the convention that sets the rules to govern autonomous machines. These four goals are accomplished in two parts of this paper. Part I offers definitions and a detailed overview of the technology used in AWS; this helps in determining whether or not AWS can fit under the current standards of LOAC rules and weapon laws. Part II argues that the autonomous weapons developed or intended for use in armed conflict must be capable of being used in accordance with LOAC. The paper discusses just war theory that includes jus in bello for the use of AWS in warfare. Further, this paper discusses the rules of distinction, proportionality, unnecessary suffering and military necessity in attack. This section is followed by discussion about prohibition on conventional weapons and customary law, which governs the law of weapons. It also evaluates the propriety of evaluating AWS by those standards, addresses potential shortcomings of such application, and proffers alternative avenues to refine current weapons laws. This paper explores that there is an urgent need for a new legal framework to regulate AWS and presents different solutions for the legal framework of AWS. The conclusion argues that a preemptive ban is irrelevant at this point and urges the appropriate authorities to develop a modern legal framework that is tailored to embrace these state-of-the-art weapons as LOAC develops.

Section snippets

Definitions

To the extent that there are no standard, universally accepted definitions of some of the key terms employed in this paper, there is need to stipulate those working definitions here since it is important to ensure the same basic understanding of those terms at the outset.²⁰ It will help in addressing the legal issues that underlie the use of AWS. For the purpose of this paper, these terms have these assigned meanings.

Autonomy of AWS

Before learning about the legal aspects and governing of AWS, it is important to discuss the relevant technology used in AWS for a better understanding of its legality. The key difference between human soldiers and AWS is that humans have the capability to distinguish between a legitimate target and a non-legitimate target. Humans use their biological senses for the distinction; whereas, AWS make the decision based on their human created technological components. This ability to decide which

The Law of Armed Conflict (LOAC)

Currently, there are no treaties that specifically address or govern the use of AWS.⁵⁵ However, there is universal consensus that the LOAC applies to AWS.⁵⁶ All warfare is governed by IHL, also known as

Weapon laws

AWS are subject to a two-part LOAC test, applicable to all weapons, under Article 51(4) and Article 36 (weapon review will be discussed in the following section) of Additional Protocol I (AP I) to the Geneva Convention.¹⁷⁵ This section of this paper is intended to help in determining if present weapon laws are sufficient

Legal challenges

The advent of AWS creates new challenges that need to be addressed.¹⁹⁹ In this section, full view of the challenges faced in the field of AWS will be addressed and organized.²⁰⁰ These challenges are organized in thematic sub-groups: Asimov's laws of Robotics, risk of cyber attack, prohibition of AWS, and positive aspects of the use of A.I. in warfare and other future challenges. The first challenge of

Positive aspects of the use of artificial intelligence in warfare

Every innovation has both negative as well as positive aspects. The biggest positive aspect of AWS is that it offers the ability to reduce the number of human lives at risk during war.³⁰⁷ The cognitive burdens on human operators will be lessened by autonomous technology, and certain functions will be

Conclusion

A.I. has great potential to benefit humanity in many ways.³¹⁹ Starting a military A.I. arms race is a bad idea, without having regulatory laws for AWS.³²⁰ Moreover, there should be a new legal framework on AWS. As a group of scientists that advocates restrictions on the use of military robots has said, “are these human-designated targets? Or are these systems automatically deciding what is a target?”³²¹