THE REGULATION OF AUTONOMOUS WEAPONS
UNDER INTERNATIONAL HUMANITARANIAN LAW:
LEGAL AND ETHICAL CONSIDERATIONS
ABSTRACT
The advent of autonomous weapon systems (AWS) presents a critical challenge to the
framework of International Humanitarian Law (IHL), which governs the conduct of
hostilities during armed conflicts. AWS, which is capable of independently selecting and
engaging targets without human intervention raise profound legal and ethical questions
about the future of warfare. This paper examines the regulatory landscape surrounding
AWS, analysing how current IHL principles such as distinction, proportionality, and
precautions in attack apply to autonomous systems. These core principles require human
judgment to assess dynamic battlefield conditions and to protect civilians from
unnecessary harm. Judgments that autonomous systems, despite technological
advancements, are currently ill-equipped to make reliably. The paper highlights the
operational deployment of semi-autonomous systems, including the 2020 use of Turkey's
Kargu-2 drone in Libya, and explores the trajectory of global military investments in AWS.
It outlines the degrees of autonomy in weapon systems, ranging from human-in-the-loop
to fully autonomous. It also discusses the legal challenges these variations pose.
Particularly problematic are accountability gaps when AWS act independently, as
responsibility becomes diffused across developers, commanders, and operators.
Additionally, the paper investigates the ethical concerns raised by delegating lethal
decision-making to machines, the erosion of human dignity, and the risk that AWS may
lower the threshold for armed conflict by distancing decision-makers from the
consequences of violence. Challenges in regulating AWS include the technological
complexity and lack of explainability in AI decision-making processes, the absence of
international consensus, and the limitations of existing legal review mechanisms such as
Article 36 of Additional Protocol I. To address these issues, the paper recommends
ensuring meaningful human control over AWS, fostering transparency and public
engagement, and developing legally binding international instruments tailored to AWS
1|Page
regulation. Ultimately, this study underscores the need to align military innovation with
the foundational principles of IHL and ethics, advocating for responsible governance to
ensure that autonomous technologies serve the protection, not the erosion, of humanity in
times of war.
INTRODUCTION
The use of weapons in armed conflicts is strictly regulated by International Humanitarian
Law (IHL). However, considering the availabilities resulting from technological
advancements, the use and regulation of autonomous weapons systems (AWS) or say, killer
robots in armed conflict remain an ongoing challenge. IHL regulates and bans the use of
different kinds of weapons in armed conflict that cause damage or injury unreasonably and
discriminatorily. Some among such weapons are laser beams causing permanent blindness,
weapons which injure by fragments undetectable in the human body through X- rays, and
such. The use of autonomous weapons is however not covered under any IHL treaties
particularly thereby, putting forward a huge matter to be addressed by nations worldwide.1
By the definition given by the International Committee of the Red Cross (ICRC) in its
analysis of the issues posed by autonomous weapons systems under IHL, an autonomous
weapon system is any weapon system with autonomy in its critical functions, that is, a
weapon system that can select (search for, detect, identify, track or select) and attack (use
force against, neutralize, damage or destroy) targets without human intervention.2 The only
human intervention required is for the launching or activation of the weapon system, after
which, it by itself performs targeting functions and attack.3 Therefore, the shift from
traditional weaponry raises profound questions about the future conduct of warfare, the
role of human judgment, and the adequacy of existing legal and ethical frameworks.
Although fully autonomous weapons have not yet been widely deployed in combat, recent
conflicts have seen the use of AI-powered munitions and drones with significant
autonomous capabilities. Notably, the Turkish-built Kargu-2 drone reportedly conducted
an attack in Libya in 2020 without a human “in the loop,” marking a milestone in the
1
Neil Davison, 'Autonomous Weapon Systems under International Humanitarian Law' (ICRC, January
2018)
[Link]
al_humanitarian_law.pdf
2
Ibid.
3
Ibid.
2|Page
operational use of such systems.4 These developments signal a trajectory toward
increasingly sophisticated and independent military platforms, with major powers
investing heavily in AWS research and deployment. The United States, for example, has
launched initiatives like the Replicator Initiative to accelerate the integration of
autonomous systems across all warfighting domains, while other nations, including China,
Russia, and Israel, are similarly advancing their autonomous military technologies.5
The integration of AWS into modern arsenals presents significant challenges for IHL,
which governs the conduct of hostilities and seeks to protect civilians and combatants from
unnecessary suffering. IHL is built upon fundamental principles such as distinction,
proportionality, and precaution, all of which require nuanced human judgment in the
application of force.6 The delegation of life-and-death decisions to machines raises
concerns about the ability of AWS to reliably distinguish between combatants and
civilians, assess proportionality in dynamic environments, and implement necessary
precautions to minimize harm. Furthermore, the diffusion of responsibility inherent in the
use of AWS brings in and complicates issues of accountability for unlawful acts, as the
legal and moral responsibilities traditionally rest with human beings.7
Beyond legal considerations, the rise of AWS leads to deep ethical debates. The removal
of human agency from lethal decision-making undermines human dignity and erodes the
moral fabric of warfare.8 There are fears that AWS could lower the threshold for armed
conflict, making war more likely by reducing the risks to one’s own forces and distancing
decision-makers from the consequences of violence.9 The unpredictability of complex
Artificial Intelligence (AI) systems further intensifies these concerns, as unintended
4
Ethan Lee, 'What are Autonomous Weapon Systems?' (Belfer Center for Science and International
Affairs, 3 February 2025) [Link]
5
Ibid.
6
International Committee of the Red Cross, International Humanitarian Law and the Challenges of
Contemporary Armed Conflicts: Building a Culture of Compliance for IHL to Protect Humanity in Today’s
and Future Conflicts (ICRC, 30 September 2024) [Link]
humanitarian-law-and-challenges-contemporary-armed-conflicts-building
7
Dr Marta Bo, Laura Bruun and Dr Vincent Boulanin, 'Retaining Human Responsibility in the
Development and Use of Autonomous Weapon Systems: On Accountability for Violations of International
Humanitarian Law Involving AWS' (SIPRI, October 2022) [Link]
8
Neil Renic and Elke Schwarz, 'Crimes of Dispassion: Autonomous Weapons and the Moral Challenge of
Systematic Killing' (2023) 37(3) Ethics & International Affairs 321
[Link]
9
'Weapons Without Witnesses: The Emerging Challenges of Autonomous Military Systems' (Legal
Vidhiya, March 2025) [Link]
autonomous-military-systems/
3|Page
actions by autonomous weapons could lead to the escalation or violations of the laws of
war.10
This paper explores the regulation of autonomous weapons under IHL by examining both
the legal and ethical concerns that are posed by such advanced technologies. By analysing
current developments, legal frameworks, and ongoing international debates, the paper aims
to contribute to a deeper understanding of how the international community can balance
technological innovation with the necessity of humanitarian protection and ethical
responsibility.
AN OVERVIEW OF THE SCOPE OF IHL
International Humanitarian Law (IHL), also called the law of armed conflict or the law of
war, is a specialized branch of international law that seeks to limit the effects of armed
conflict for humanitarian purposes. Its primary aim is to protect individuals who are not
participating in hostilities, such as civilians, medical and religious personnel, and those
who are wounded or have surrendered. IHL also imposes restrictions on the means and
methods of warfare, prohibiting weapons and tactics that cause unnecessary suffering or
superfluous injury.11
The scope of IHL extends to both international armed conflicts (between states) and non-
international armed conflicts (within a state, involving state and non-state actors). It is
applicable only during times of armed conflict and does not address the legality of going
to war (jus ad bellum), but rather focuses on the conduct within war (jus in bello). The
main sources of IHL are international treaties, especially the four Geneva Conventions of
1949 and their Additional Protocols, as well as customary international law and general
legal principles.12 Key principles underpinning IHL include the obligation to distinguish
between combatants and civilians, the requirement of proportionality in attacks, and the
necessity to use only the force required to achieve legitimate military objectives. All parties
to a conflict must treat protected persons humanely and without discrimination.
10
Roman V Yampolskiy, 'Unpredictability of AI: On the Impossibility of Accurately Predicting All Actions
of a Smarter Agent' (2020) 7(1) Journal of Artificial Intelligence and Consciousness 109
[Link]
11
International Committee of the Red Cross, What is International Humanitarian Law? (ICRC, 2021)
[Link]
12
Michael Shriney, 'International Humanitarian Law' (iPleaders, 11 June 2022)
[Link]
4|Page
Implementation and enforcement of IHL are primarily the responsibility of states, with
oversight by organizations such as the International Committee of the Red Cross (ICRC).13
In May 2025, nations convened at the United Nations to discuss the regulation of AI-
powered weapons, acknowledging the urgent need for binding international standards.14
Despite these discussions, consensus remains elusive, with key states like the U.S., Russia,
China, and India favouring national guidelines over global treaties. This ongoing debate
underscores the challenges of adapting IHL to rapidly advancing military technologies.
IHL establishes the legal framework governing conduct during armed conflicts, including
the use of weapons. Several fundamental principles are particularly relevant to autonomous
weapon systems:
1. Distinction: Parties to a conflict must distinguish between civilians and
combatants, and between civilian objects and military objectives at all times.15
2. Proportionality: Attacks that may cause incidental civilian casualties or damage
to civilian objects that would be excessive compared to the anticipated military
advantage are prohibited.16
3. Precautions in Attack: Those who plan or decide upon attacks must take all
feasible precautions to avoid or minimize civilian harm, including cancelling or
suspending attacks if circumstances change.17
4. The Martens Clause: This historical provision links ethical considerations to legal
requirements by stipulating that in cases not covered by existing treaties, civilians
and combatants remain under the protection of "the principles of humanity" and
"the dictates of public conscience".18
13
Elisa Massimino, 'How is IHL Implemented?' (ICRC Commentary on International Law, 14 August
2017) [Link]
14
'Nations Meet at UN for 'Killer Robot' Talks as Regulation Lags' (Neuron Expert, 12 May 2025)
[Link]
15
Supra note 1.
16
Ibid.
17
Ibid.
18
Ibid.
5|Page
DEGREES OF AUTONOMY AND HUMAN CONTROL IN AWS
The level of autonomy in weapon systems can be conceptualized along a spectrum defined
primarily by human involvement in the targeting process. A widely referenced model
categorizes AWS based on the level of human control:
1. Human-in-the-Loop Weapons: Systems that can select targets but only deliver
force with explicit human command.19
2. Human-on-the-Loop Weapons: Systems that can select targets and apply force
under human oversight, with operators retaining the ability to override system
decisions.20
3. Human-out-of-the-Loop Weapons: Systems capable of selecting targets and
delivering force without any human input or interaction after initial activation.21
LEGAL CHALLENGES OF AUTONOMOUS WEAPONS UNDER IHL
The regulation of AWS under IHL brings into consideration various legal concerns, which
include:
Challenges to the Principle of Distinction. One of the fundamental challenges posed by
autonomous weapon systems is their ability to reliably distinguish between civilians and
combatants in complex operational environments. This distinction is not merely a matter
of visual or sensor-based identification but requires contextual understanding and often
subjective judgment. When an autonomous system uses sensors to match detected shapes
against a "target profile," questions arise about whether the system can accurately identify
if the shape triggering a strike is indeed a military vehicle rather than a civilian car.22 The
technical capacity of machines to make such nuanced distinctions, particularly in dynamic
conflict environments, remains a significant concern despite advances in artificial
intelligence and sensor technology.
19
Christian Alwardt and Martin Krüger, Autonomy of Weapon Systems (IFAR, February 2016)
[Link]
20
Ibid.
21
Ibid.
22
Neil Davison, 'What You Need to Know About Autonomous Weapons' (ICRC, 26 July 2022)
[Link]
6|Page
Proportionality Assessment Difficulties. The rule of proportionality requires a complex
weighing of anticipated military advantage against expected civilian harm – a distinctly
human judgment involving both factual and value assessments. Autonomous systems face
significant limitations in conducting such evaluations, which require understanding of
military strategy, tactical situations, and the valuation of human life against military
objectives. An autonomous weapon system operating independently for extended periods
may be unable to conduct the ongoing proportionality assessments required by IHL,
particularly as circumstances change during operations.23 This limitation becomes more
acute when systems operate in dynamic environments where civilian presence and
activities may fluctuate.
Precautions in Attack and Accountability Concerns. IHL requires combatants to take
all feasible precautions to avoid or minimize harm to civilians, including cancelling attacks
when circumstances change. The implementation of such precautions becomes
problematic with autonomous systems that operate independently without continuous
human supervision. As noted by the ICRC, if "a mobile autonomous weapon system
searches for targets over a wide area and for a long duration, without human supervision
and communication, the commander who authorized the launch of the weapon and the
operator who activated it will not know exactly where and when an attack will take
place".24 This creates significant challenges for implementing the precautionary measures
required under IHL. Furthermore, autonomous weapons may create accountability gaps.
While IHL places responsibility on human commanders and operators, determining who is
legally accountable for unintended civilian harm becomes increasingly complex as the
causal chain between human decisions and weapon effects becomes more attenuated.
ETHICAL CONSIDERATIONS –
Human Dignity and Machine Decision-Making. Beyond legal considerations,
autonomous weapons raise profound ethical questions about delegating life-and-death
decisions to machines. Many argue that allowing machines to make decisions about when
to use lethal force against humans fundamentally violates human dignity and the moral
requirement that such decisions involve human moral agency and judgment. This concern
23
Supra note 1.
24
Ibid.
7|Page
connects directly to the Martens Clause's reference to "principles of humanity" and
"dictates of public conscience," suggesting that growing public discomfort with
autonomous weapons may itself have legal significance under IHL.25 The question
becomes whether societies are willing to accept critical decisions about the application of
force being made by machines rather than humans with moral agency.
Risk of Lowering the Threshold for Armed Conflict. A significant ethical concern is
that autonomous weapons may lower the political and psychological barriers to initiating
armed conflict. By reducing the risk to one's own forces, autonomous systems could make
the decision to engage in armed conflict seemingly less costly, potentially increasing the
frequency of conflicts and undermining global security. The separation between human
decision-makers and the battlefield, already expanded through remote warfare
technologies, would be further increased by autonomous systems. This distancing effect
raises concerns about warfare becoming more abstract and detached from its human
consequences for those deciding to initiate or continue hostilities.
Delegation of Life-and-Death Decisions. The delegation of lethal decision-making to
machines raises concerns about the erosion of human dignity and moral agency. The ICRC
emphasizes that decisions to kill should remain under human control, as machines lack the
capacity for empathy and moral reasoning. The Vatican has also called for a halt in the
development of AWS, advocating for international agreements to ensure human oversight
in lethal decisions.26
Control and Predictability Problems. The complexity of modern artificial intelligence
systems raises concerns about predictability and control. As systems become more
sophisticated, particularly those employing machine learning algorithms, their decision-
making processes may become less transparent and less predictable to human operators,
creating risks of unintended consequences that could include violations of IHL or
unintended escalation of conflicts. This unpredictability presents a fundamental challenge
to commanders' ability to ensure that weapon systems will function in accordance with
25
Supra note 1.
26
Vatican, Antiqua et Nova: Note on the Relationship Between Artificial Intelligence and Human
Intelligence (28 January 2025)
[Link]
nova_en.html
8|Page
their intentions and within the bounds of legal and ethical constraints in real-world conflict
environments.
CHALLENGES IN REGULATING AUTONOMOUS WEAPONS
Even while the regulation of AWS is attempted, it hinders or makes such regulation
difficult due to challenges which include:
Technological Complexity and Explainability. AWS often operate using complex
algorithms and machine learning, leading to the "black box" problem, where their decision-
making processes are opaque and difficult to interpret. This lack of transparency hampers
the ability to assess compliance with IHL and to investigate potential violations. Ensuring
explainability and traceability in AWS operations is essential for accountability and legal
scrutiny.
International Consensus and Regulatory Efforts. Efforts to establish international
regulations for AWS have faced challenges due to differing national interests and strategic
considerations. While some countries advocate for legally binding agreements, others
prefer non-binding guidelines. The United Nations has been a platform for discussions, but
consensus remains elusive. The ICRC urges states to adopt new legally binding rules to
address the risks posed by AWS.27
Article 36 Reviews and Legal Compliance. Article 36 of Additional Protocol I to the
Geneva Conventions mandates that states review new weapons to ensure compliance with
international law.28 However, the dynamic nature of AWS, particularly those employing
machine learning, complicates these reviews. Regular and rigorous assessments are
necessary to ensure that AWS operate within legal boundaries throughout their
deployment.
RECOMMENDATIONS FOR REGULATION
To uphold legal and ethical standards, it is imperative to ensure meaningful human control
over AWS. This involves human oversight in the selection and engagement of targets, as
27
International Committee of the Red Cross, ICRC Position on Autonomous Weapon Systems (12 May
2021) [Link]
28
Additional Protocol (I) to the Geneva Conventions, Article 36 (1977) [Link]
treaties/api-1977/article-36
9|Page
well as the ability to intervene and deactivate the system if necessary. Such control
mechanisms are vital to maintain accountability and moral responsibility.
The international community should work towards the development of legally binding
instruments specifically addressing AWS. These instruments should define the permissible
scope of autonomy in weapon systems, establish accountability frameworks, and enforce
compliance with IHL. Human Rights Watch advocates for a pre-emptive ban on fully
autonomous weapons to prevent potential violations of international law.
Governments and developers should promote transparency in the development and
deployment of AWS. Engaging the public in discussions about the ethical implications of
AWS can foster informed decision-making and societal consensus. Public input is crucial
in shaping policies that reflect collective moral values and uphold human dignity.
CONCLUSION
The development of autonomous weapon systems presents unprecedented challenges to
the international legal framework governing armed conflict. While technological
innovation continues to advance military capabilities, the fundamental principles of IHL
and ethical considerations demand careful attention to how these technologies are
developed and deployed.
The international community faces the task of developing regulatory frameworks that
account for rapid technological change while preserving the humanitarian protections
established by IHL. This requires ongoing dialogue among states, international
organizations, technical experts, and civil society to ensure that advances in autonomous
weapons technology do not outpace the legal and ethical frameworks designed to limit the
suffering caused by armed conflict.
The challenge ahead is to ensure that human judgment, responsibility, and accountability
remain central to decisions about the use of force, even as weapon systems become
increasingly autonomous. By maintaining appropriate human control over weapons, the
international community can work to ensure that technological developments serve to
enhance rather than undermine the protection of humanity during armed conflict. The legal
and ethical questions surrounding autonomous weapons will remain at the forefront of
international humanitarian discourse as these technologies continue to evolve.
10 | P a g e