An Agenda for Action: Alternative Processes for Negotiating a Killer Robots Treaty

After close to 10 years of talk with no tangible results, countries should find an effective forum to negotiate a treaty on autonomous weapons systems. The vast majority of states parties to the Convention on Conventional Weapons (CCW) agree that a new legally binding instrument is urgently needed to address the serious threats posed by this emerging technology. Nevertheless, they have not taken concrete steps toward making this goal a reality.

Meetings of the CCW, which began to address autonomous weapons systems in 2014, have failed to produce even voluntary commitments on these novel and dangerous weapons systems, much less a binding instrument. It is time for states to pursue action elsewhere. This report calls on states to initiate a treaty-making process based on past humanitarian disarmament models and explains why such a process would be a more efficient and effective response to the myriad of concerns raised by autonomous weapons systems.

As Human Rights Watch and others have repeatedly pointed out, autonomous weapons systems present a host of ethical, moral, legal, accountability, and security challenges. For example, delegating life-and-death decisions to machines would dehumanize armed violence and pose a threat to human dignity. Autonomous weapons systems that operate without meaningful human control would face difficulties complying with international humanitarian and human rights law due their lack of human qualities, such as judgment and compassion. They would also create a gap in criminal and civil accountability, and the development of such a new technology could lead to an arms race and proliferation to actors with little regard for international law.

Because of these concerns, as well as the increasingly rapid pace of the systems’ development, more than 70 states, civil society organizations, and the International Committee of the Red Cross (ICRC) regard a new treaty with prohibitions and restrictions as necessary, urgent, and achievable. Such a treaty should include (1) prohibitions on autonomous weapons systems that inherently lack meaningful human control, (2) prohibitions on autonomous weapons systems that target people, and (3) regulations, or positive obligations, on all other autonomous weapons systems to ensure meaningful human control.

As Part I of this report highlights, progress on a legally binding instrument on autonomous weapons systems is unlikely to occur under the CCW’s auspices. The forum is hampered by the lack of a common purpose among all states parties, a reliance on consensus for decision-making, the absence of deadlines, and threats to inclusivity. These four factors have interfered with the goals of most states parties, and the Sixth Review Conference in 2021 and the Group of Governmental Experts (GGE) meetings in December 2021, March 2022, and July 2022 accomplished little. Proposals from states parties advocating for a binding instrument, despite representing the majority view, were repeatedly blocked by major military powers, which took advantage of the consensus process. Given recent history, it is unlikely that the CCW meetings in November 2022 and beyond will take a different course.

Rather than accepting continuing stagnation under the CCW, proponents of a legally binding instrument should seek an alternative forum to negotiate a new treaty. Part II of this report examines the processes that successfully led to the Mine Ban Treaty, the Convention on Cluster Munitions, and the Treaty on the Prohibition of Nuclear Weapons. The first two treaties were negotiated in independent processes undertaken outside of United Nations auspices, but with the support of UN agencies and officials. The third was initiated by the UN General Assembly (UNGA). All three processes emerged from states’ frustration with a consensus-based forum and ended up producing a strong treaty in a timely fashion.

Part III identifies four characteristics that allowed these alternative processes to succeed. States were united by a common purpose, which prevented fundamental differences from hijacking talks and allowed participants to focus on details early on. Decisions were made by voting, rather than consensus, which created stronger and more equitable results and prevented powerful states from blocking results. States set clear and ambitious deadlines that ensured the process progressed and concluded efficiently. A commitment to inclusivity, especially in the independent processes, enhanced negotiations by engaging the views of a wide range of states, civil society and international organizations, survivors, and other experts. Proponents of a legally binding instrument on autonomous weapons systems should recognize the benefits of these alternative processes and use them as models for their own treaty process.

Finally, Part IV of this report recognizes and responds to concerns that states may have about proceeding with a treaty through an alternative process. Even proponents of a legally binding instrument may raise questions about the appropriateness of leaving the CCW track and the legitimacy, participants, and scope of an independent or UNGA-initiated process. The success of earlier alternative treaty processes, combined with the unlikelihood of progress under the CCW, demonstrates that following the lead of disarmament precedent is the best way to move forward with negotiations on autonomous weapons systems.

As the ninth year of CCW meetings on the topic comes to a close, the discussions have failed to produce the much-needed prohibitions and regulations on autonomous weapons systems that a majority of states desire. With major military powers getting ever closer to developing these dangerous systems, alternative options need to be pursued. It is time for states to initiate a process elsewhere to negotiate a new treaty on autonomous weapons systems.

Current diplomatic talks on autonomous weapons systems are going nowhere and show that the CCW forum is incapable of producing a credible outcome on this issue. Human Rights Watch and the International Human Rights Clinic (IHRC) at Harvard Law School, therefore, call on states to pursue an alternative process to negotiate a legally binding instrument to address rising concerns over autonomous weapons as soon as possible. Such a process should have the following characteristics:

1.       A common purpose of creating new law that ensures all negotiating parties are committed to work in good faith to achieve their shared goal;

2.      A model of decision-making by majority vote, as opposed to consensus, for all decisions of procedure and substance;

3.      A clear and ambitious deadline for the negotiation and adoption of the treaty, so that the instrument can urgently respond to technological developments while also providing sufficient time and resources for the consideration of humanitarian and security concerns; and

4.      An inclusive approach to ensure wide participation from a range of states, UN agencies, international institutions, and nongovernmental organizations.

 

Over the past year, states parties to the CCW have convened on multiple occasions to discuss what to do about concerns over “lethal autonomous weapons systems,” including at the convention’s Sixth Review Conference in December 2021 and at Group of Governmental Experts meetings held in December 2021, March 2022, and July 2022. Instead of advancing concrete actions, however, these meetings have repeatedly demonstrated that the CCW forum is incapable of achieving meaningful progress on this issue. It has been hampered by the failure to achieve a common purpose across all states parties, a consensus-based approach to decision-making, the absence of deadlines despite the urgency of the problem, and threats to inclusivity. These factors continue to plague the CCW process and have played a large role in the body’s inability to advance regulations of autonomous weapons systems.

Overview: December 2021 – July 2022

In December 2021, CCW states parties held their Sixth Review Conference at the UN in Geneva, to discuss, among other matters, lethal autonomous weapons systems and to consider proposals for new CCW protocols. The vast majority of states parties called for a legally binding instrument on autonomous weapons systems. For example, the Non-Aligned Movement, representing more than 100 states, highlighted the “urgent need” for a legally binding instrument. Austria called for “urgent negotiations on prohibitions.” The Holy See urged the CCW to adopt “an ambitious forward-looking approach” through the commencement of negotiations on a new instrument. New Zealand’s disarmament minister appealed for negotiations to commence on legally binding prohibitions and controls. South Africa proposed the CCW “move into the phase of concrete negotiations” on new law.

Despite overwhelming support for a strong mandate for 2022, the Review Conference failed to deliver. CCW meetings operate by consensus and minority actors such as Israel, Russia, the United Kingdom, and the United States, opposed a binding protocol. As a result, the Review Conference ultimately adopted a mandate for the GGE to “consider proposals and elaborate, by consensus, possible measures” regarding lethal autonomous weapons systems. A large number of states parties expressed deep dissatisfaction at this weak outcome. By its conclusion, the December conference demonstrated that rather than providing a multilateral forum in which states can collaborate to address the challenges posed by autonomous weapons systems, the CCW more often results in gridlock on this topic.

The December, March, and July GGE meetings were likewise defined by gridlock. In December, states parties failed to reach consensus on any recommendations for the Review Conference. In March, Russia stalled the first two days of meetings and singlehandedly put a stop to formal discussions. Throughout the July meetings, Russia continued its obstructionist actions, including by challenging civil society participation in formal discussions. While July’s substantive conversations again revealed convergence around a legally binding instrument with prohibitions and regulations, the final draft of the GGE report lacked information on the details of a potential agreement. The GGE recommended simply rolling over its mandate from December 2021 and left unresolved the length of next year’s session. States parties to the CCW will decide next year’s mandate when they meet in November 2022.

Lack of a Common Purpose

Progress under the CCW has largely stalled because, even though the positions of most states parties have coalesced around the need for a legally binding instrument on autonomous weapons systems, there is no common purpose uniting all states parties. After nine years of discussions, a minority of countries disagree about the purpose of the enterprise and the best way to move forward. While most states concur that some combination of legally binding prohibitions and regulations are necessary, a small number of the world’s major military powers oppose such restrictions, arguing instead for voluntary commitments, such as manuals or codes of best practices. Russia poses further challenges to these discussions as it opposes any new rules or regulations. While including a diverse range of opinions initially sparked interesting debate, the split between the shared goal of most states and the divergent goals of a few powerful countries has rendered recent discussions unproductive.

Written submissions to and discussions at the July 2022 meeting showed that most CCW states parties support some version of a legally binding solution with a combination of prohibitions and regulations (the so-called “two-track approach”). Eight states submitted a proposal for a new CCW protocol with prohibitions on autonomous weapons systems that violate international humanitarian law or are designed to be used outside meaningful human control, and with regulations to prevent data bias and environmental harm while ensuring accountability and meaningful human control. A working paper from seven European states called for prohibiting weapons that absolutely lack human control, while regulating other weapons to ensure compliance with international humanitarian law, accountability, and human control. Peru, Norway, Belgium, New Zealand, and Bulgaria, among others, have recognized the popularity of the two-track approach. Italy, Austria, and the ICRC likewise voiced support for the two-track approach, with Italy noting that such an approach maintained an element of human control, while not inhibiting the advancement of technology.

The July meeting also demonstrated that the divergent views championed by a few countries pose serious challenges to the majority. Despite the widespread agreement discussed above, states that insist on voluntary commitments or are entirely against new rules and regulations hinder further progress in the CCW forum. Furthermore, Russia and other states developing autonomous weapons systems repeatedly highlighted the purported benefits of these weapons systems, even as an overwhelming number of states agreed on the need for regulations or outright bans. On the first day of meetings, Russia touted the “advantages” of autonomous weapons systems as a response to the “emotional” discussions about the weapons’ risks. Despite widespread condemnation, Russia, joined by India and the US, argued on the second day that autonomous weapons can reduce civilian casualties and ensure human control. Russia claimed that such weapons cannot “panic” like humans. Once again, these comments were met with swift condemnation, with Cuba noting that humans’ tendency to feel “panic” may be a positive attribute, and Panama stressing that the forum is not meant for arms promotion.

While the vast majority of CCW states share the common goal of regulating or prohibiting autonomous weapons systems, the presence of states advocating for the very same weapons prevents progress toward new international law. Ultimately, it is unreasonable to place hope in a disarmament venue that hosts a set of actors with perspectives that fundamentally cannot be reconciled with the perspective shared by the majority. As discussed below, the CCW’s reliance on consensus and absence of deadlines exacerbate this challenge.

Consensus-Based Decision-Making

Thus far, the primary difficulty to making progress under the CCW is that its states parties employ a consensus approach to decision-making. A 2005 UN legal opinion defines consensus as “the absence of objection rather than a particular majority.” Practically, consensus means that a single state can reject a proposal, even if every other state agrees to it, which leads to the lowest common denominator among all parties. In the CCW context, the consensus model has allowed minority states such as Russia and the US to block the majority’s proposals for regulating and prohibiting autonomous weapons systems. The CCW meetings over the past year exemplify how the consensus-model has stymied popular and necessary action on autonomous weapons systems.

For example, despite dedicating significant time to drafting a report, the GGE was unable to reach consensus on any recommendations for the CCW’s Sixth Review Conference in December 2021. As discussed above, the Review Conference, which met the following week, adopted only a weak discussion mandate. The vast majority of states sought far stronger language than simply to “consider proposals,” and many voiced dissatisfaction with the result. A group of 13 like-minded states advocating for a treaty on autonomous weapons systems expressed their frustration over the incomplete work, and Palestine, one of its members, added in its national capacity that delaying action on autonomous weapons systems placed national interests over the interests of humanity. Switzerland and 15 other states were similarly dismayed and made a joint statement on the last day of the 2021 Review Conference that said:

Simply repeating the discussions we have already had at previous GGEs does not do justice to the work already completed or the task at hand. Nor should the consensus practice of this forum be abused. Those who would invoke it to protect their positions should do so responsibly and seldom, and only when all other avenues are exhausted.

The March 2022 GGE meeting reinforced the CCW’s inability to advance meaningful discussion under the consensus model. The first two days were stalled by Russia, which refused to adopt the agenda or allow substantive discussions because its representatives could not travel to Geneva ostensibly due to sanctions imposed following the country’s full-scale invasion of Ukraine that February. Because the CCW operates by consensus, Russia’s refusal to take up agenda items singlehandedly put a stop to formal discussions, forcing the meetings to move to informal mode for the remaining three days. Russia’s unilateral delay was costly: two of the allotted five days were spent discussing procedural matters, and the remaining three were spent behind closed doors in informal sessions that lacked procedural significance.

The July 2022 meetings were likewise marked by Russian obstructionism and an overall lack of progress due to the consensus approach. On the fourth day of meetings, Russia attempted to include in the draft report that there were no objections to March’s meeting being held in informal mode, despite the fact that, thanks to its abuse of consensus, the group had spent two full days disputing this procedural matter. Russia also asked that the report specify that the group only discussed substantive issues on July 25 and 26, even though the GGE had had substantive discussions in March and throughout the July meetings. After states parties debated the issue at length, the chair decided to hold informal consultations later that night to resolve the disagreement. Once again, the consensus model both wasted valuable time and stopped meaningful debate on autonomous weapons systems.

Russia’s tactics have raised questions about the legitimacy of the CCW as a diplomatic forum. The CCW’s reliance on consensus has also allowed a persistent few to slow progress and undermine a convergence of views supporting a two-track approach. At best, the body will produce a voluntary commitment that fails adequately to address the dangers posed by autonomous weapons systems. At worst, the CCW will fail to adopt even voluntary guidelines on these novel and dangerous weapons.

Absence of Deadlines

Because CCW states parties have set no deadlines for their work, progress is likely to continue to be slow. This general lack of urgency and accountability has proven very damaging. For instance, the 2022 GGE could not agree to recommend a mandate that was notably different from the one the group had this year. Even though most states wanted the mandate to reflect the group’s convergence around a two-track approach to a legally binding instrument, a small subset of states was able to block the majority’s proposal with no consequence.

By the close of the July meetings, states parties were also unable to agree on how many days the GGE will meet next year. While most states supported Germany’s proposal that the GGE meet for 20 days in 2023, Russia and a few other states dissented. As a result, the issue of whether to meet for 20 or 10 days remains unresolved, and will need to be debated again at the annual CCW meeting scheduled for November 2022. The proposal to decrease the number of days from 20 to 10 suggests that this forum could produce even less work. Absent significant deadlines, there is no end in sight for the cyclical stagnation that has defined the past nine years of discussions.

Threats to Inclusivity

Recent CCW meetings have also posed threats to the inclusivity of disarmament. As discussed above, the consensus process has favored major military powers who are more likely for both political and strategic reasons to individually block consensus. As a result, it has been difficult for small and medium-sized states to enact change under the CCW, despite representing the majority view.

In addition, there have been threats to the inclusion of civil society and international organizations. Despite long-standing precedent for including these groups in CCW meetings, at the July meeting, Russia and India both attempted to block nongovernmental organizations (NGOs) from participating in discussions of the draft report. These countries interrupted one NGO as it was speaking, arguing that civil society was not allowed to comment on the text being debated by states parties. Many other states, including Austria, Belgium, Canada, Chile, France, Germany, Ireland, Mexico, Montenegro, New Zealand, the US, and others, quickly came to the NGO’s defense. The chair ruled that the NGO could continue to speak.

On the last day of the July discussions, Russia for a second time tried to exclude civil society. Once again, many states came to the support of international organizations and civil society, and the chair upheld the ruling from the prior day. Russia’s opposition to civil society’s contributions was not only an attempt to exclude a valuable segment of participants, but also a way to slow down the meeting’s progress and challenge the forum’s legitimacy.

The Answer to Gridlock

Given the gridlock over the course of these meetings, it is hard to imagine how CCW states parties could adopt a protocol that imposes any meaningful obligations on the international community. Despite widespread support for a legally binding instrument with prohibitions and regulations, the body lacks a common purpose to make that end result a reality. A minority of states have been able to abuse the consensus model repeatedly by treating it as a grant of veto power. The lack of clear and ambitious deadlines has allowed ongoing delays in dealing with an urgent issue. Threats to inclusivity challenge the transparency and substantive quality of discussions. These issues raise questions about the feasibility, efficacy, and legitimacy of any outcome in the CCW, which will at best result in a watered-down voluntary commitment.

Given that states parties have provided no indication that they will ever negotiate or adopt a legally binding instrument on autonomous weapons systems within the CCW, states should pursue an alternative process, looking at models to those that were used for other humanitarian disarmament treaties. This report examines such alternatives, identifies their advantages, and shows why they are appropriate in the autonomous weapons systems context.

 

Over the past 25 years, disarmament has achieved great success through processes outside the CCW, the Conference on Disarmament, and other disarmament bodies of the UN. The two main alternatives for the negotiation of weapons treaties are independent (also known as standalone) state-led processes and UN General Assembly-initiated processes. The independent Ottawa and Oslo Processes were used to negotiate the 1997 Mine Ban Treaty and the 2008 Convention on Cluster Munitions, respectively, while a UNGA-initiated process produced the 2017 Treaty on the Prohibition of Nuclear Weapons.

In each case, the processes were created in response to flawed consensus-based forums and proved to be more efficient, inclusive, and successful. The processes exemplify the humanitarian disarmament approach to governing weapons, which seeks to address arms-inflicted human suffering through the establishment and implementation of norms. Given their similar origins and purposes, the independent and UNGA-initiated processes should inform future negotiations of a treaty on autonomous weapons systems.

Mine Ban Treaty

The Ottawa Process created an innovative approach that has served as a model for future treaty negotiations. The process began in the wake of the CCW’s flawed First Review Conference held in Vienna in 1994 and then Geneva in 1995.States participated in the Review Conference at a time when awareness was increasing over the humanitarian consequences of widespread use of antipersonnel landmines, particularly in conflicts in Afghanistan, Angola, and Cambodia. By the time of the CCW’s First Review Conference, the Office of the UN High Commissioner for Refugees (UNHCR), UN Children’s Fund (UNICEF), UN Secretary-General Boutros Boutros-Ghali, and the ICRC, among others, were advocating for a total ban on the use of weapons, spurred on by the International Campaign to Ban Landmines (ICBL).

The CCW’s First Review Conference, however, adopted only an amended protocol on landmines (Amended Protocol II), which fell far short of a ban and was instead “overly complex and insufficiently stringent to deal with the extent of the humanitarian crisis” posed by landmines. One author attributed the failure to structural issues inherent to the CCW, particularly the consensus-based decision-making system.

Frustrated with the Review Conference’s lack of a credible outcome, 75 pro-ban states gathered in Ottawa, Canada, from October 3–5, 1996 with the goal of developing a strategy to achieve a worldwide ban on antipersonnel landmines. At the close of the conference, Lloyd Axworthy, the Canadian foreign affairs minister, challenged those present to return to Ottawa at the end of 1997 to sign a treaty to ban landmines. Axworthy’s announcement launched the intensive Ottawa Process, successfully culminating in the adoption of the Mine Ban Treaty in Oslo, Norway, on September 18, 1997.

The Ottawa Process followed an ambitious 14-month timeline that provided states with multiple opportunities to influence the evolving treaty text. The conferences were held in the first half of 1997 in Vienna in February, Bonn in April, Johannesburg in May, and Brussels in June before the diplomatic negotiations were held in Oslo on September 1–18. A total of 121 states signed the treaty when it opened for signature at a high-level conference in Ottawa from December 3-4.

The structure of the Ottawa Process facilitated the conclusion of a strong treaty in an efficient and inclusive manner. A core group of states set commitment mechanisms in place at two stages to ensure that states participated in good faith and with a common goal. Participation in the initial October 1996 conference in Ottawa was limited to states that had endorsed a draft declaration calling for the urgent negotiations of a treaty to ban antipersonnel landmines. Furthermore, to participate fully in the final text negotiations, states were required to sign the Brussels Declaration, which committed states to negotiate a treaty by the end of 1997. States that did not sign the Brussels Declaration could attend the Oslo negotiations as observers, as could UN agencies, the ICRC, and the ICBL. Unlike the CCW, the Oslo negotiations operated according to a two-thirds majority voting rule for substantive changes to the draft treaty text. The rule, which was never put to the test, was modelled after the rules of procedure for the 1977 Geneva Convention Additional Protocols.

The Ottawa Process benefitted from the leadership and participation of a geographically diverse range of countries. It was led primarily by Canada and a core group of states, including Austria, Belgium, Germany, Ireland, Mexico, the Netherlands, Norway, the Philippines, South Africa, and Switzerland. With 120 states involved in the final negotiations in Oslo, the Ottawa Process engaged both mine producers and affected states.

NGOs played an instrumental role in the Ottawa Process as evidenced by the awarding of the 1997 Nobel Peace Prize to the ICBL and its then-coordinator Jody Williams for starting an initiative “which in the space of a few years changed a ban on anti-personnel mines from a vision to a feasible reality.” The ICBL was the first to frame the landmine issue in humanitarian terms and call for an international ban on the weapon. It pushed states to negotiate and continues to promote and monitor the Mine Ban Treaty’s universalization and implementation to this day. Campaigners, including landmine survivors and military veterans, were vocal advocates for a strong and comprehensive treaty and were present throughout the negotiations, both inside and outside the conference hall. One US negotiator reportedly disdainfully described the negotiations as “a new diplomacy in which NGOs had virtually equal status inside the delegations and you had NGO delegations sitting around the table. It was a free for all.” At the Oslo Conference, the ICBL and its members successfully lobbied to include provisions on victim assistance in the final text and to counter US proposals to insert a geographic exception, definitional loopholes, and reservations aimed at preventing swift entry into force.

The Mine Ban Treaty, which entered into force on March 1, 1999, ultimately proved to be much more successful than its predecessor, CCW Amended Protocol II. The Mine Ban Treaty provides an absolute prohibition on the use, stockpiling, production, and transfer of antipersonnel landmines as well as on assistance with those banned activities. It requires states to destroy their stockpiles within four years and to clear mined areas within 10 years.

The treaty directly has led to the clearance of hundreds of square kilometers of mine-affected land, the destruction of more than 55 million antipersonnel mines, and a significant decrease in the number of new mine victims each year. While 33 states, including China, Russia, and the US, have not joined the Mine Ban Treaty, it has become an indispensable tool in stigmatizing antipersonnel mines and provides the only comprehensive international framework for their complete eradication.

Convention on Cluster Munitions

The Oslo Process, which resulted in the adoption of the Convention on Cluster Munitions, closely followed the Ottawa Process model. It followed years of failure by the CCW to heed calls to address the humanitarian impacts of cluster munitions.

In 2003, the CCW adopted Protocol V on Explosive Remnants of War (ERW), with the aim of providing post-conflict measures applicable to all weapons and highlighting the importance of clearing ERW. The protocol ignored, however, growing concerns over cluster munitions, including how to prevent their wide area effects at the time of attack and their deadly explosive remnants left afterwards.

Norway championed the issue, initiating the Oslo Process after the CCW’s Third Review Conference in 2006 failed to commit to negotiate a legally binding instrument to address the unacceptable harm posed by cluster munitions. The Third Review Conference took place only months after Israel’s widespread and devastating use of cluster munitions in Lebanon, which underscored the human cost of the weapons. Before the conference even started, concerned states parties submitted a proposal that called for negotiating a specific protocol on cluster munitions, but several states opposed the proposal. As the CCW follows a consensus approach to decision-making, the rejection by these states meant that the Review Conference would not address cluster munitions. Norway announced on the last day of the conference that it would launch the Oslo Process.

Norway sought help from a core group of states that actively called for treaty negotiations during the CCW process and later led the Oslo Process itself: Austria, the Holy See, Ireland, Mexico, New Zealand, and Peru. The group’s activities included hosting conferences and drafting discussion and treaty texts.

The Oslo Process took place at five conferences spread over the course of 15 months. At the initial Oslo Conference in February 2007, 46 states endorsed the Oslo Declaration on Cluster Munitions, which set the goal of adopting a legally binding instrument to ban cluster munitions and establish stockpile destruction, clearance, and victim assistance obligations by the end of 2008.Conferences held in Lima in May 2007, Vienna in December 2007, and Wellington in February 2008 saw more than 100 states converge to consider key elements needed in the draft treaty. States that endorsed the Wellington Declaration, which reiterated the Oslo Declaration’s goals and committed them to negotiate a new treaty based on a draft text, were allowed to participate in the final negotiations in Dublin in May 2008. The 107 states at the Dublin conference adopted the Convention on Cluster Munitions by acclamation on May 30.

The success of the Oslo Process was due in large part to its leadership and ways of working. The core group drafted the discussion and treaty texts and incorporated suggestions and recommendations after each conference.The process avoided using the “bracketed texts” of the UN system that present all options, including the weakest ones, in brackets or italics. This practice prevented prolonged debates and unnecessary delays. States participating in the Dublin negotiations also agreed to rules of procedure that employed the two-thirds majority voting rule for matters of substance, such as revisions to adoption of text. The rule ensured that negotiations could run smoothly and that a minority of states could not weaken the treaty or halt progress unilaterally. The president of the final Dublin Diplomatic Conference was empowered to put the whole treaty to a vote if agreement could not be reached, but the convention was adopted without the need to call one.

The Oslo Process was also characterized by its inclusiveness. The core group was geographically diverse, and participating states included stockpilers and producers of cluster munitions as well as countries contaminated by their remnants, such as Laos and Lebanon. International organizations were also instrumental in garnering support and strengthening the legitimacy of the process. The UN Mine Action Service along with other mine clearance organizations provided technical expertise. The UN Development Program (UNDP) helped administer Norway’s sponsorship program for resource-constrained states. The ICRC distributed legal analyses of the treaty text, made frequent interventions to critique the text through a humanitarian lens, and provided recommendations for amendments.

The central involvement of NGOs grouped under the banner of the Cluster Munition Coalition and the participation of dozens of cluster munition and landmine survivors known as “ban advocates” proved instrumental in keeping the humanitarian imperative for the treaty at the heart of the process. Civil society participated throughout the negotiations at a status virtually equal to that of states, except that during the negotiations themselves they could not submit formal language proposals or adopt the final treaty. Campaigners undertook a variety of activities, such as highlighting the humanitarian impact of cluster munitions, providing legal analyses of proposed text, making statements in plenary sessions and working groups, and advocating behind the scenes. Despite some major military powers, including China, Russia, and the US, opting out, the Oslo Process is widely lauded as an inclusive effort where different stakeholders from diverse geographical locations, survivors, and dedicated humanitarian civil society actors, could participate in meaningful ways.

The Oslo Process resulted in a convention that comprehensively bans the use, production, stockpiling, and transfer of cluster munitions as well as assistance with those prohibited activities. The treaty also obligates states parties to destroy their stockpiles within eight years and to clear areas contaminated by cluster munition remnants within 10 years. States are required to provide medical care, rehabilitation, and psychological support to victims as well as provide for their social and economic inclusion.

The convention’s 110 states parties and 13 signatories have made significant progress in fulfilling these obligations. There have been no reports of cluster munition use by any state party since the convention’s adoption. To date, states parties have destroyed 99 percent of total declared stockpiles, destroying nearly 1.5 million cluster munitions and 178 million submunitions. The Convention on Cluster Munitions has also strengthened norms stigmatizing the weapons, and cluster munitions are being phased out of service in many of the 47 countries with stocks that have not joined the convention.

Treaty on the Prohibition of Nuclear Weapons

The 2017 Treaty on the Prohibition of Nuclear Weapons was adopted through a UN General Assembly-initiated process, rather than an independent one, but it emerged from the similar failure of an existing treaty regime that relied on consensus-based decision-making. More specifically, the Nuclear Non-Proliferation Treaty, which aimed to prevent the spread of nuclear weapons technology and to achieve nuclear disarmament, became associated with two overwhelming problems: (1) the failure of states to abide by their obligations to pursue a treaty on “general and complete disarmament under strict and effective international control,” and (2) the simultaneous expansion and modernization of existing nuclear arsenals.

Motivated by the successful humanitarian disarmament framing of the Convention on Cluster Munitions, non-nuclear-armed states expressed comparable concern over the “catastrophic humanitarian consequences of any use of nuclear weapons” at the 2010 Nuclear Non-Proliferation Treaty Review Conference. This sentiment gave rise to the “Humanitarian Initiative,” a series of conferences focused on achieving total nuclear disarmament and mitigating harms from nuclear weapons use and testing. The Humanitarian Initiative was led by conference hosts Norway, Mexico, and Austria with the intimate involvement of civil society and international organizations.

In stark contrast to the gradual “step-by-step”path to disarmament advocated by nuclear-armed states favoring the Nuclear Non-Proliferation Treaty status quo, the Humanitarian Initiative conferences, held from 2013-2014, emphasized the catastrophic consequences of nuclear weapons use and testing and called on nuclear-armed states to intensify their efforts at disarmament. The Humanitarian Initiative centered its attention on the impact of nuclear weapons on people and the environment, relying on the testimony of victims and other experts. At the third and final international conference in Vienna, 127 states signed the “Humanitarian Pledge,” which called on states parties to the Nuclear Non-Proliferation Treaty to renew their commitment to adopting a legally binding instrument, to take measures to reduce the risk of harm from nuclear detonations, and to work with all relevant stakeholders including international organizations and civil society, all with the goal of prohibiting and eliminating nuclear weapons.

In the wake of renewed enthusiasm in favor of nuclear disarmament and a failure to achieve consensus on a final declaration at the 2015 Non-Proliferation Treaty Review Conference, 138 UNGA members voted in favor of establishing an open-ended working group to develop laws and norms for achieving and maintaining a nuclear-weapon-free world. Following the working group’s recommendation to draft a legally binding instrument, 123 states in 2016 voted in favor of a UNGA resolution that convened a “United Nations conference to negotiate a legally binding instrument to prohibit nuclear weapons, leading towards their total elimination.”

Subsequent to the passage of the resolution, the first week-long session at the UN in New York in March 2017 involved expert presentations and interactive discussions. Delegations from 132 states as well as civil society and international organizations elaborated on, developed, or altered their formal positions.

The second session, held from June–July 2017 in New York, involved several open plenary sessions to which civil society organizations were invited, as well as closed-door negotiations on the treaty text that were closed to civil society. A core group comprised of Austria, Brazil, Ireland, Mexico, Nigeria, and South Africa played a critical role in leading regional discussions, ultimately driving consensus among neighboring countries. Although the negotiating session in June–July took place over only three weeks, the time pressure was alleviated by the fact that delegations in attendance were mostly aligned on the general objective and scope of the treaty, whereas opponents, notably the nuclear-armed states, declined to join the negotiations.

From the first plenary session, the conference was marked with relatively active involvement by civil society and international organizations, though the UN process placed more restrictions on their engagement than the independent processes. Notable among the organizations were the International Campaign to Abolish Nuclear Weapons (ICAN) and the ICRC. Civil society and international organizations played a key role in emphasizing humanitarian goals early in the process through participation in the 2013–2014 international conferences and, during the negotiation stage, in plenary sessions and side events. Furthermore, victims and affected individuals had a role in sharing their experiences and steering the text toward particular framings. Neither civil society organizations nor the ICRC were permitted to join the closed meetings negotiating the final text of the treaty, however.

Like the Ottawa and Oslo Processes, the negotiations on the Treaty on the Prohibition of Nuclear Weapons used a two-thirds majority voting rule for substantive issues. At the conclusion of the negotiations, the voting rule ensured that the treaty was adopted despite one abstention and one vote against. While the treaty was nearly universally accepted by parties, its adoption would have been blocked by a single opposing state under a consensus approach.

The Treaty on the Prohibition of Nuclear Weapons contrasts with the Nuclear Non-Proliferation Treaty by explicitly prohibiting the development, testing, production, stockpiling, transfer, use, and threat of use of nuclear weapons. Instead of creating different obligations for nuclear and non-nuclear armed states in a manner upholding the status quo, the Treaty on the Prohibition of Nuclear Weapons absolutely bans nuclear weapons for all states parties. It does so with the humanitarian consequences of nuclear weapons as an explicit motivating force.

The treaty also foregrounds the lived realities of victims of nuclear weapons use and testing, much like the Mine Ban Treaty and the Convention on Cluster Munitions. Similar to Article 5 of the latter, Article 6 of the Treaty on the Prohibition of Nuclear Weapons obligates states to “adequately provide age- and gender-sensitive assistance, without discrimination, including medical care, rehabilitation, and psychological support, as well as provide for their social and economic inclusion.” The treaty imposes an affirmative duty on states to engage in environmental remediation in areas contaminated as a result of nuclear weapons use or testing. Like other humanitarian disarmament treaties, the Treaty on the Prohibition of Nuclear Weapons obligates states to cooperate with, and assist each other in its implementation. As of October 2022, the treaty has 68 states parties and an additional 26 signatories. It has strengthened political and legal norms against nuclear weapons’ development, stockpile, and use as well as started to set new standards for victim assistance and environmental remediation.

 

The case studies discussed above illuminate four elements of treaty processes that are particularly conducive to achieving strong humanitarian disarmament instruments in a timely fashion—a common purpose, voting-based decision-making, clear and ambitious deadlines, and a commitment to inclusivity. These characteristics, whether realized in an independent or UNGA-initiated process, would also address many of the shortcomings of the CCW. By heeding the lessons of past processes, proponents of a legally binding instrument on autonomous weapons systems could free themselves of the CCW stalemate and pursue a more effective and efficient approach to a new treaty.  

Common Purpose

In all three case studies, state participants had a common purpose from the beginning of the treaty process. The fact that they set a clear goal for themselves enabled collective action and attention to specific issues without disagreement over the big picture. A common purpose also facilitated agreement on process, which in turn benefited substantive discussions toward stronger and more comprehensive provisions. In other words, adopting a shared vision at the outset allowed for a more efficient negotiating process because it sharpened the deliberations of the parties and facilitated the removal of traditional procedural hurdles. An alternative process focused on autonomous weapons systems would similarly benefit from the early commitment to a common purpose by allowing those states that want stronger norms to set the tone, facilitating practical choices that streamline the process, and clearly communicating that all negotiating partners are operating in good faith toward the same goal.

The three case studies demonstrate the method and value of establishing a common purpose from the outset. Both the Ottawa and Oslo Processes required states to endorse a declaration at their first meeting pledging to pursue a ban treaty. States later had to reiterate their commitment to such an instrument to participate in the formal negotiating conferences. The common purpose behind the Treaty on the Prohibition of Nuclear Weapons negotiations was expressed in the UNGA resolution that launched the process. According to an Austrian ambassador actively involved in that process, the “high unity of purpose among the negotiating States” was “[o]ne of the key reasons” why the negotiations reached their goal so quickly.

An alternative process would allow states with the common purpose to prohibit and regulate autonomous weapons systems to work together to negotiate a new legally binding instrument. After nearly a decade of discussions in the CCW, most countries’ positions have “coalesced around the need for a new legal instrument including prohibitions and regulations.” Committed states should take advantage of this momentum by moving to an independent or UNGA-initiated forum that would help them achieve that goal. States that are not fully dedicated to this end can disengage and, if they choose, come on board at a later date. An alternative process with a common purpose that promotes ownership and good faith among participating states would allow negotiations to proceed without major disagreements, disruptions, or unnecessary delays.

Voting-Based Decision-Making

The decision-making rules of the alternative processes discussed in this report make for stronger and more equitable outcomes. The three case studies demonstrate how use and abuse of the consensus rule under the CCW and Nuclear Non-Proliferation Treaty have inhibited substantive progress, let alone the initiation of negotiations on new legally binding instruments. Under a consensus model, a single state’s ability to interfere with the will of the majority leads to compromises that represent the lowest common denominator. In addition, more powerful states can generally veto proposals they oppose without hesitation, while less influential states may be ignored when they speak out in the same manner or feel constrained in their ability to block consensus because doing so would alienate important allies. For this reason, a consensus approach often privileges the positions and interests of the most powerful states, even if they are in the minority on a particular issue. In 2010, former Canadian foreign affairs minister Lloyd Axworthy noted, in the context of Amended Protocol II, that CCW disarmament negotiations were “held hostage to the position of the most recalcitrant members” due to the consensus model. As they strive to regulate autonomous weapons, states ought to learn from these past experiences and move forward with a model more conducive to effective regulation.

A decision-making rule that allows for voting leads to stronger results. All three case studies discussed in this report emerged from failed consensus processes, i.e., under the CCW or the Nuclear Non-Proliferation Treaty. When turning to alternative processes, states rejected the consensus approach and chose to adopt a two-thirds majority voting rule for matters of substance and simple majority rule for procedural matters, which have long been common rules of procedure for multilateral treaty negotiations. Although states can strive for consensus during discussions, the possibility of a vote incentivizes states to remain flexible and open to compromise. The independent processes of the Mine Ban Treaty and Convention on Cluster Munitions never came to a vote, but the prospect of one encouraged states to work together to develop a satisfactory result with majority support, rather than be forced to accept the weakest option. As discussed above, the existence of the Treaty on the Prohibition of Nuclear Weapons depended on the fact that it could be adopted by a two-thirds majority vote rather than requiring consensus when the Netherlands challenged the final proposed text.

Alternative processes have produced outcomes that would have likely been impossible to achieve under a consensus rule. Where the CCW failed to produce a ban on antipersonnel mines despite a clear humanitarian need and a demonstrated willingness by most states, and where it failed to produce any regulations whatsoever on cluster munitions, the independent processes produced strong and effective bans for states parties and ultimately shaped international norms even for states not parties. Furthermore, where the Nuclear Non-Proliferation Treaty created a division between nuclear “haves” and “have-nots” and failed to significantly advance nuclear disarmament, the Treaty on the Prohibition of Nuclear Weapons produced absolute prohibitions and established that nuclear weapons are illegal for all states parties.

As autonomous weapon technology continues to emerge, developing treaty law outside of the CCW would be an opportunity to set strong obligations and norms whose strength would reside, in part, in their near universal acceptance. As long as states remain in the consensus-bound CCW, there is unlikely to be a feasible pathway to a legally binding instrument despite the significant convergence of a majority of states’ positions. Only by moving to a voting-based procedure can this majority of states break free from the willful opposition of the minority and act on their desire to craft a treaty that prohibits and regulates autonomous weapon systems.

Clear and Ambitious Deadlines

Alternative processes are also characterized by clear and ambitious deadlines. These deadlines increased the speed and efficiency of the negotiations of the three instruments discussed above and supported the rapid development of strong treaty law.

In each case study, deadlines set early in the treaty process ensured the negotiations continued to move along and were concluded in a timely manner. They were announced in the declarations adopted at the start of the Ottawa and Oslo Processes and in the UNGA resolution that called for negotiations of a nuclear weapons ban treaty. The deadlines led to processes that ranged from seven to 15 months from initiation to treaty adoption, an efficient pace in international law. Ultimately, the quality of the results showed that states did not have to sacrifice robust and comprehensive provisions in the name of speedy outcomes.

Given how rapidly autonomous weapons technology is developing, it is crucial that the international community develop treaty law to regulate these weapons as soon as possible. Although discussions on autonomous weapons have been ongoing in CCW meetings since 2014, negotiations on a new protocol have yet to begin. An alternative forum, whether independent or UNGA-initiated, would empower the international community to set ambitious deadlines and act quickly to ensure that the range of concerns associated with these technologies are addressed prior to their widespread development and proliferation.

Commitment to Inclusivity

The inclusion of a wide range of states and non-state actors—civil society and international organizations—is another factor common to the case studies that bolstered the credibility and comprehensiveness of the three treaties discussed in this report. Alternative processes can amplify the voices of a wide range of participants with different interests, concerns, and areas of expertise, ensuring that many aspects of the issues pertaining to the weapons system at issue are properly discussed and negotiated.

The legitimacy of the alternative processes in all three case studies depended in large part on the geographic and other types of diversity of the states involved. An inclusive approach ensured participation of states from all regions of the world and provided a platform for small and medium-sized states to address the humanitarian consequences of specific types of weapons. Affected states and, in the case of an independent processes, users and producers also participated. The absence of certain major military powers, such as China, Russia, and, except in the Ottawa Process, the US, shifted the focus away from their concerns and toward those of the majority.

In addition to facilitating the participation of a diverse range of states, alternative processes also create heightened opportunities for the involvement of and collaboration with non-state actors. The insights, expertise, and relationships that civil society and international organizations bring to the table are highly valuable for the development of effective treaty language and the process of building momentum towards strong regulation.

As discussed above, in all three case studies, civil society and international organizations helped bring states to the negotiating table by investigating and publicizing the humanitarian impacts of arms. Along with survivors, they also played a key role during the treaty processes by lending expertise and diverse perspectives, which informed the final treaty language. During the Treaty on the Prohibition of Nuclear Weapons negotiations, however, the level of inclusion of civil society and international organizations was more mixed. While these groups were able to engage in the plenary and public events in the margins of the negotiation conference, they were excluded from closed-door meetings during which the finalization of the treaty language occurred, a serious omission and a concerning downside of any UNGA-initiated process that follows that model.

Moving to an alternative process for autonomous weapons systems that ensures inclusivity during the negotiation of a treaty on autonomous weapons systems would have many benefits. Given that autonomous weapons systems have been described, with concern, as the next revolution in warfare, their development, production, and use would affect every state. Therefore, it is critical that the voices of a diverse range of states are heard during the negotiating process. In addition, as the three case studies demonstrate, civil society and international organizations can increase public support for a new treaty and help develop treaty obligations that reflect key humanitarian concerns. Their participation also lays the groundwork for further work to ensure strong implementation of and compliance with treaties. The complex and ever-evolving nature of this emerging field of technology means that the negotiations would benefit from a wide range of legal, ethical, technical, and security expertise. The inclusion of civil society experts, such as members of the Stop Killer Robots campaign, in an alternative process on autonomous weapons systems would expand the pool of resources available to states when drafting the terms of treaty provisions and ensure that outcomes reflect the views of those with specialized knowledge and experiences. It would also help promote the humanitarian goals of disarmament.

 

Despite the numerous advantages of independent and UNGA-initiated processes, some states have expressed strident opposition to alternative pathways. Even certain proponents of a legally binding instrument on autonomous weapons systems may hesitate to negotiate that instrument outside the auspices of the CCW, where the issue has been addressed for so long. For example, some states may argue they are obligated to continue engaging in CCW discussions regarding proposals for regulating autonomous weapons systems, believe that the CCW framework has greater legitimacy because it is part of the UN, want to ensure the involvement of major military powers in negotiations, or fear that security issues will be overlooked in an alternative treaty process. This part responds to each of these concerns in turn and concludes that none of them should stand in the way of negotiating a new treaty in an independent or UNGA-initiated process.

Prioritization of Goal over Forum

Some CCW states parties have privately expressed concern that pursuing an alternative treaty process at this point would be seen as interfering with progress in the CCW context, in contravention of what they have long been advocating for. The introduction in March 2022 of three proposals to address the challenges raised by autonomous weapons systems heightened their fears of appearing to be acting in bad faith.

States that support a comprehensive legally binding instrument prohibiting and regulating autonomous weapons systems should not, however, lose sight of their shared goal. Two of the proposals tabled at the March GGE fall short of a legally binding instrument. A group of six states led by the US proposed a set of “principles and best practices” to strengthen international humanitarian law compliance, while the UK called for development of a manual stating how international humanitarian law applies to lethal autonomous weapons systems. Although such non-binding documents might help shape and standardize military and state practice, they would legitimize the use and further development of autonomous weapons systems, so long as those activities abided by the agreed-upon guidelines. The third proposal was for a CCW protocol with prohibitions and regulations, but in the current environment, the adoption of such an instrument is unrealistic.

Proponents of a new legally binding instrument on autonomous weapons systems should recognize that their end goal is more important than the forum in which it is achieved. Pursuit of a new treaty outside of the CCW does not prevent ongoing work under the CCW. Furthermore, CCW states parties agreed on Guiding Principles of 2019 that find “[t]he CCW offers an appropriate framework for dealing with the issue,” which indicates that it is not the only such framework. Treaty proponents should therefore focus their efforts on the process that promises the best results and not be restricted by a flawed one.

Legitimacy of an Independent Process

Some states may be concerned that an autonomous weapons systems treaty negotiated outside the UN would face issues of legitimacy. Having not emerged from the UN framework, such a treaty might have less symbolic and actual authority than a CCW protocol or a UN General Assembly-initiated treaty.

Precedent from the Mine Ban Treaty and the Convention on Cluster Munitions shows these concerns are unfounded. Although the independent Ottawa and Oslo Processes took place outside the UN, UN agencies played an active role during the negotiations. For example, deminers brought field experience that illuminated the immediate and long-term humanitarian consequences of the weapons.

In addition, the UN continues to be deeply involved in the administration and monitoring of both treaties produced by independent processes. The Mine Ban Treaty and the Convention on Cluster Munitions require initial and regular reporting to the UN secretary-general. The secretary-general also acts as the depositary for both treaties, convenes meetings of states parties and review conferences, oversees amendments and withdrawals, and handles requests for clarification regarding matters of compliance. UN actors have the ability to attend all meetings of states parties and review conferences as observers. The UN system plays an integral role in compliance and cooperation for both treaties.

The widespread ratification of the Mine Ban Treaty and the Convention on Cluster Munitions have demonstrated that independent procedures are on par with or even more accepted by the international community than the CCW. At the time of writing, the CCW had 126 high contracting parties, with its five protocols ranging from 96 to 119 parties. The Mine Ban Treaty had 164 states parties, and the Convention on Cluster Munitions had 110 states parties. Despite some claims that the latter two treaties are not legitimate or at the same “level” as the CCW or UN-originated instruments, state practice and universalization levels have proven otherwise.

Inclusion of and Ability to Influence Users and Developers

Some states have contended that the CCW is the best forum for addressing autonomous weapons systems because all the major military powers can participate. Given the high-tech nature of autonomous weapons systems, it can be valuable to have the participation of and support from states that are developers and potential users of these systems.

An alternative treaty process, whether independent or UNGA-initiated, would be open to participation from all states (not just CCW states parties). It would differ from the CCW forum in that states would have to commit to a common purpose to ensure that progress could be made. In addition, it would not be constrained by consensus, and thus would prevent certain states from having undue influence over others.

Users and producers of the weapons being addressed by negotiations joined the independent treaty processes discussed in this report. For example, more than half of the world’s cluster munitions stockpilers and two-thirds of its producers and users participated in the Oslo Process. Although the nuclear-armed states did not participate in the negotiations of the Treaty on the Prohibition of Nuclear Weapons, their absence eliminated opposition to ambitious treaty provisions, and 124 other states engaged making for robust negotiations that produced a strong and widely supported result. States favoring a treaty on autonomous weapons systems should therefore not be daunted if users and developers decline to participate in the negotiations.

Regardless, adopting a treaty is the best way to create standards on autonomous weapons systems that can influence states not party to the agreement. A strong treaty, negotiated through an alternative process, would significantly enhance international norms by stigmatizing the development, production, and use of autonomous weapons systems that are inherently legally or morally unacceptable and setting affirmative standards for requiring meaningful human control over all other systems that rely on sensor processing to select and engage targets.

Addressing Security Issues

The CCW is often described as a place that balances security and humanitarian concerns. Some states may therefore be concerned that an alternative process will neglect security concerns as a new treaty is negotiated.

These states need not worry, however, because the security concerns raised by autonomous weapons systems are widely recognized, including by states that favor a new legally binding instrument. The development of such weapons systems could give rise to an arms race, and mass production of such weapons at economies of scale is “virtually inevitable.” Proliferation to irresponsible states or non-state armed groups could lead to violence against civilians and regional instability. In addition, as Sri Lanka cautioned during a GGE meeting in 2015, autonomous weapons systems have the potential to “escalate the pace of warfare,… [to] undermine the existing arms controls and regulations, to aggravate the dangers of asymmetric warfare, and [to] destabilize regional and global security.”

A new legally binding instrument is the only way to adequately address these security concerns. It would provide more specific and enforceable rules than a voluntary agreement. In addition, many of the security risks detailed above—an arms race, proliferation, and lowering the threshold for war—stem not only from the potential use of autonomous weapons systems, but also from their development and production. While international humanitarian law focuses on the use of weapons systems, a legally binding instrument could strengthen this body of law by establishing clear prohibitions and regulations not only on the use of autonomous weapons systems, but also on their development and production.

Given that a new CCW protocol is unlikely, state parties should look for an alternative process to address the security as well as the moral, legal, accountability, and other concerns raised by autonomous weapon systems. The longer states stay in the CCW forum, the longer developers of autonomous weapons systems have to hone new technologies and achieve commercialization. CCW gridlock merely delays possible responses to security and humanitarian threats and paves the way for an arms race, proliferation, and other security threats along with civilian suffering.

States should, therefore, pursue an independent or UNGA-initiated process to negotiate a treaty that prohibits and regulates autonomous weapons systems as soon as possible.

 

This report was authored and edited by Bonnie Docherty, senior researcher in the Arms Division of Human Rights Watch. She is also the associate director of armed conflict and civilian protection and a lecturer on law at Harvard Law School’s International Human Rights Clinic (IHRC). Ryen Bani-Hashemi, Madeleine Cavanagh, Arnaaz Ameer, Alexa Santry, Elliot Serbin, Rosalinn Zahau, and Lisa Wang, students in IHRC, made major contributions to the research, analysis, and writing of this report. Steve Goose, director of the Arms Division, and Mary Wareham, advocacy director of the Arms Division, edited the report. James Ross, legal and policy director, and Fred Abrahams, associate director for program, provided legal and programmatic reviews respectively.

Production assistance was provided by Susan Aboeid, associate with the Arms division. The report was prepared for publication by Travis Carr, Publications Officer. Kathleen Rose, senior editor, reviewed the press release accompanying the report.

 

https://www.hrw.org/report/2022/11/10/agenda-action/alternative-processes-negotiating-killer-robots-treaty