Autonomous Weapon Systems: Reframing and Communicating AI Ethics

Scenario addressed and perspective taken

This article goes over the ethical dimensions of Lethal Autonomous Weapons Systems (LAWS) from the perspective of a representative of the Australian Human Rights Commission. As a representative of organisations that work towards banning these weapons, I argue that LAWS fundamentally violates human dignity while creating gaps in accountability. It threatens to transform warfare in ways that go against core humanitarian principles. This paper goes over all the core ethical problems of implementing LAWS worldwide and demonstrates how meaningful human control must be maintained over lethal decisions.

1. Introduction

LAWS is an autonomous weapons system that can independently select, engage, and inflict lethal damage to any target, without further human intervention. Over 40 nations have called for bans or moratoria on LAWS development so far. The UN Secretary-General António Guterres declared them “politically unacceptable and morally repugnant” (Wikipedia contributors, 2025). Additionally, evidence of their deployment in conflicts in Ukraine and Libya goes to show that this is an ongoing issue and not a future threat (Santoni de Sio, F., & van den Hoven, J, 2018).

On the other hand, the defence industry argument states that activating autonomous weapons systems is equivalent to “pulling a trigger”. But this ignores the nature of moral agency and responsibility. Although there is an argument claiming that the person pressing the “on” button bears full accountability, in reality, this ignores the black box nature of these lethal systems (Guo, J, 2025). This fundamental disconnect between human decision-makers and lethal outcomes is discussed as “a responsibility gap”.

Assumptions

1. Human dignity is a non-negotiable value in warfare.
2. Moral responsibility cannot be delegated to machines.
3. Existing international humanitarian laws require human control over any lethal force decisions.
4. Weapon development must adhere to precautionary principles whenever fundamental rights are at risk.

2. Core Ethical Problems

2.1. Human Oversight and Moral Agency

LAWS create a system with a responsibility gap between the programmers, commanders and the system itself. The ethical problem arises when we don’t know who bears responsibility when there’s a case where the system inflicts lethal damage on civilians due to some algorithm bias or sensor failure. The programmer cannot anticipate all the operation situations, the commander can always disclaim responsibility for machine decisions, and the manufacturers can always evade trouble by calling it a technical complexity (Guo, J, 2025).

This represents the fundamental violation of the principles of moral agency, the capacity to make ethical decisions based on right or wrong and be accountable for those decisions (Physiopedia, 2025). It is “the ability to bring impartial moral principles to bear on the determination of what, all considered, morally ought to be done” (Johnson L. S. M, 2021). No matter how we look at it, and regardless of the sophistication, a machine will always lack consciousness and the capacity for moral reasoning. Moral agents can only have responsibilities, but they themselves cannot bear responsibility or accountability over their actions.

Causes

Unlike predictable systems that are preprogrammed to execute certain commands under certain circumstances, LAWS is a technology that uses the sensory data, target and an algorithm to decide by itself, the lethal action it takes (ICRC, 2018). In other words, the problem arises when machines, that lack consciousness and the capability to make moral decisions, step in to mediate in the situation. This creates a gap between human reasoning and lethal outcomes that cannot be bridged by simply claiming “activator pulls the trigger”.

Stakeholders and impact

The accountability gap affects multiple stakeholders

1. Victims and their families – Cannot identify who will be responsible for their harm.
2. Military personnel – They can face ambiguous responsibilities that can undermine military ethics.
3. Programmers – Face passive responsibility designing machines that cannot fully control.
4. Society – Loses the ability to maintain peace and control legitimate violence.

Alignment with ACS

Value 1.1: The Primacy of the Public Interest (ACS, 2023) – LAWS creates a situation where civilians can be harmed without clear accountability. 

2.2. Violation of Human Dignity

Ethical problem

Being killed by an autonomous machine fundamentally violates human dignity, since it reduces the person to just a mere data point to be processed.  The repulsion people usually feel for LAWS comes from the idea that machines making life-and-death decisions crosses the fundamental moral line (Campaign to Stop Killer Robots, 2013).

Causation

LAWS killing a person can strip the victim of fundamental human dignity. The cause that violates human dignity stems from the fact that machines do not comprehend the value of a human life. Machines can never look at a human life as anything more than a pattern from sensory data. Human rights advocates call this “digital dehumanisation”. It matters not just if a person is killed or not, but also how they are killed, and the decision that is being taken (Campaign to Stop Killer Robots, 2013). 

Alignment with ACS

Value 1.3: Enhancement of Quality of Life (ACS, 2023) – Humans have a right to retain dignity even in conflict, and the way LAWS work goes against it.

2.3. Normalisation of Autonomous Killing

Ethical problem

Normalisation of deviance describes how organisations gradually start normalising abnormal behaviours as standard practice (Vaughan D, 1996). When this is applied to LAWS, we can see how this technology will gradually make its way from military to domestic law enforcement. When these lethal systems designed for foreign battlegrounds are adapted to domestic law enforcement and crowd control, they also bring their dehumanised approach to assessing threats. 

Causation

In addition to normalisation of deviance, existing adopted methods paint a convincing picture of how normalization of LAWS in the future is very likely. Australia’s state police force has already established certain operations groups that are trained by the Australian Defence Force using military tactics and weapon systems (Antony Funnell, 2019). Without clear legal boundaries, it is not impossible for LAWS to be deployed against Australian citizens in a way that violates civil liberties.

Alignment with ACS

ACS Code of Professional Conduct – Value 1.1 and Value 1.3 (ACS, 2023) – The ACS demands that professionals put public interest and rights first (ACS, 2023). Additionally, it requires tech to improve people’s lives, not create fear or oppression. Deploying a dehumanising threat assessment against Australian citizens would directly violate these codes of ethics.

3. Recommendations

Recommendation 1: International Moratorium During Treaty Negotiation

While we pursue a permanent ban or safer implementation, an immediate moratorium on LAWS deployment would be necessary. This moratorium can remain in place until a comprehensive international treaty is negotiated.  The scope of this treaty should cover the production and deployment of such autonomous weapons systems.

This international moratorium should only be lifted when

1. A robust international treaty with proper verification mechanisms is put in place.
2. Technical solutions to the accountability gap are decided and verified.
3. Adequate laws, safeguards and legal boundaries are put in place against misuse of this technology.

Recommendation 2: Clarify Chain of Responsibility

As discussed in the ethical problem “human oversight and moral agency”, the very nature of LAWS creates a responsibility gap and blurs the line when it comes to who is accountable. Clear legal frameworks must establish who bears responsibility when things go wrong. 

1. Designers – Hold AI developers and manufacturers accountable for foreseeable harms from their side (Santoni de Sio, F., & van den Hoven, J., 2018). This is because designers and AI developers carry the direct responsibility to anticipate and minimise, since in this case, they are shaping systems that would harm or end lives. Santoni de Sio & van den Hoven (2018) argue that developers are responsible for making sure that any avoidable risks from the technology they build must be avoided, especially when the stakes are high, like in LAWS
   
2. Commanders – Maintain the traditional military command accountability while making sure that the commander is not liable for unpredictable autonomous decisions. The reason for this is that commanders only make the decision to deploy LAWS, but the actual harm is targeted and caused by the autonomous machines. The current framework of accountability might tend to hold the commanders accountable for actions they have no control over. This can only hinder us from enforcing accountability on people who are actually accountable for the killings. The traditional command responsibility collapses for this reason; therefore, we must reinforce the framework to adapt to LAWS.
   
3. Organisations that build LAWS – Hold organisations accountable for developing and selling LAWS weapons. Unlike developers who design systems and commanders who deploy these systems, organisations that build LAWS operate at a scale where their decisions shape the market. The chain of responsibility must hold these organisations the most accountable, since they decide to build these systems that take human life without direct human control. These organisations must be legally obliged to practice rigorous safety assessments, disclose any faults in the system, and restrict deployment when the system cannot guarantee compliance with codes of ethics and international humanitarian laws. By establishing this specific line of accountability, by not letting the blame fall on the decision-makers or the designers, we can prevent unethical implementation and the accountability gap discussed before.
   
4. State – Hold nations accountable under international humanitarian law for LAWS deployment regardless of specific individual accountability.

The implementation of such a chain of responsibilities requires amendments to domestic criminal laws and international humanitarian laws. The principles of passive responsibility, where individuals and organisations must be held accountable for all the harms they failed to prevent, must be implemented specifically for LAWS (Technische Universiteit Delft, 2024). 

Recommendation 3: Strengthen Professional Codes

The Australian Computer Society (ACS) Code of Professional Ethics provides us with a great foundation, but it must be strengthened by aligning with the LAWS technology, where certain codes of ethics are established to regulate the use of technologies like LAWS in an ethical way. The goal of new codes must be to restrict the unnecessary use, production, and testing of lethal weapons. 

Some recommended codes that can be adapted to the ACS Code of Professional Ethics can look like

1. Participation in LAWS must strictly contribute to meaningful human control and development – A code like this can prevent the unethical normalisation of the LAWS technology. As discussed before, without clear boundaries, it is likely that LAWS will make their way from military to domestic law enforcement. This code of ethics can contribute to creating such a boundary.
   
2. Ethics training for members working in the defence and security sector, especially in projects involving autonomous lethal weapons, is mandatory – Ethics training can have a positive impact on how LAWS are being implemented and used. Keeping the members aware means reducing the risk of unethical use of the technology
   
3. “Whistleblower Protection” for members who refuse to work on LAWS or report concerning or unethical developments (James O’Toole, 2013). Articles on whistleblowing reveal that there is a fundamental tension between loyalty towards employers and fairness towards potential victims. The psychology of whistleblowing suggests that by emphasising “larger loyalty” towards the public over the employers can encourage ethical reporting (Dungan, J., Waytz, A., & Young, L., 2015).

Reference List

1. Wikipedia contributors. (2025, August 31). Campaign to Stop Killer Robots. In Wikipedia, The Free Encyclopedia. Retrieved 12:18, November 12, 2025, from https://en.wikipedia.org/w/index.php?title=Campaign_to_Stop_Killer_Robots&oldid=1308794621
2. Santoni de Sio, F., & van den Hoven, J. (2018). Meaningful Human Control over Autonomous Systems: A Philosophical Account. Frontiers in robotics and AI, 5, 15. https://doi.org/10.3389/frobt.2018.00015
3. Guo, J. (2025). The ethical legitimacy of autonomous Weapons systems: reconfiguring war accountability in the age of artificial Intelligence. Ethics & Global Politics, 18(3), 27–39. https://doi.org/10.1080/16544951.2025.2540131
4. Physiopedia. (2025, March 31). Moral Agency. https://www.physio-pedia.com/index.php?title=Moral_Agency&oldid=367474
5. Johnson L. S. M. (2021). Shifting the Moral Burden: Expanding Moral Status and Moral Agency. Health and human rights, 23(2), 63–73.
6. International Committee of the Red Cross (ICRC). (2018). Ethics and autonomous weapon systems: An ethical basis for human control?  ​​
7. Campaign to Stop Killer Robots. (2013, April 23). Urgent action needed to ban fully autonomous weapons: Non-governmental organisations convene to launch Campaign to Stop Killer Robots [Launch statement]. https://www.stopkillerrobots.org/wp-content/uploads/2013/04/KRC_LaunchStatement_23Apr2013.pdf
8. Australian Computer Society. (2023). Code of Professional Ethics
9. Vaughan, D. (1996). The Challenger launch decision: Risky technology, culture, and deviance at NASA. University of Chicago Press.
10. Antony Funnell. (2019). Why the creeping militarisation of our police has experts worried. https://www.abc.net.au/news/2019-09-20/creeping-militarisation-of-police-why-experts-are-concerned/11517266
11. Technische Universiteit Delft. (2024) Summary of Ethics: Active & Passive Responsibility Concepts in Ethics
12. James O’Toole. (2013). The Whistle-Blower’s Quandary: Fairness or Loyalty. https://www.scu.edu/ethics/focus-areas/business-ethics/resources/the-whistle-blowers-quandary-fairness-or-loyalty/
13. Dungan, J., Waytz, A., & Young, L. (2015). The psychology of whistleblowing. Current Opinion in Psychology, 6, 129-133. https://doi.org/10.1016/j.copsyc.2015.07.005