2nd Meeting - 1st Session Group of Governmental Experts on Lethal Autonomous Weapons Systems 2026

Good, let's start again. I have one more high contracting party taking the floor for the first time, and that's Sierra Leone. And then we move to, again, to the list of observers. But let me first pass the floor to distinguished representative of Sierra Leone. You have the floor.

Thank you very much, Chair. Thank you. As this is the first time we are addressing the session, we would like to make a general comment on the status of our work under the current mandate before we provide feedback on box 1. Thank you for your indulgence. CILoE welcomes the Chair's revised rolling text and appreciates the inclusive approach taken to incorporate constructive contributions from States. As it currently stands, the text contains the structure and substance of a set of elements that we view as a good basis for negotiations on a legally binding instrument to regulate autonomous weapons systems. As we work to complete the GSE mandate this year, we believe states have now advanced their common understandings sufficiently enough. Despite some loss of useful text from time from previous versions text, The rolling text remains a good basis for formulating by consensus, as called for in our mandate, elements that can serve as a foundation for future negotiations on a legally binding instrument. Box 1, Characterization, Overall Assessment. He already welcomes the progress in Box 1, particularly constructive compromise reflected in paragraph 1, addressing the scope of "dithal" and the essential clarification in paragraph 1. CLU supports this provision as essential to preventing the characterization from being interpreted too narrowly. However, we believe the current formulation does not fully address concerns about the inclusion of identify, as highlighted by previous speakers. There's a risk it will be interpreted too narrowly and not reflect other common human inputs prior to activation, such as setting generalized target profiles, a list of potential targets, or operating parameters. In this regard, we support the ICRC's suggested options to strengthen this provision by, A, clarifying that identify refers to the AWS's matching potential targets against a generalized target profile, ensuring other forms of human input prior to activation do not exclude systems of characterization, and, B, expanding the clause to states, and I quote, The fact that the human operator sets the list of potential targets, the generalized target profile or the operational parameters of the system does not exclude the system from decharacterization." End of quote. Without such clarification, the characterization excludes some systems that require regulation. On paragraph 1B, Sierra Leone has consistently maintained that the term "lethal" should be deleted, as it has no basis in international humanitarian law. However, we welcome paragraph 1 as a constructive compromise that clarifies the term and encompasses injury and damage to objects, not just loss of life. If "lethal" must be retained for consensus, this clarification is absolutely essential. I thank you, Chair, and I thank you, all colleagues. Thank you.

Let me thank the distinguished representative of Sierra Leone for sharing the positions of his delegation and for responding to proposals made by others. Next on my list is the distinguished representative of Article 36. You have the floor.

Thank you, Chair, and good afternoon, everyone. I wanted to engage specifically on box 1a, on the issue of identify, and perhaps to try to engage a little bit with some of the comments that have already been made in the substantive conversation this morning. I think that, from our side, we have not previously been in favour of including identify. I did not see it as necessary to include it. because of a concern that it can narrow the definition or the characterisation that we are working with here, to narrow it unnecessarily. However, we also recognise that this is an issue where there is a process at work of trying to bring the group together on a shared understanding of what it really is that we are trying to capture here. I just wanted to note that it was very helpful this morning, with a few different interventions, I noted particularly from the USA and from India where there was some narrative sort of description in those interventions to further explain how the systems were understood. And I think those kind of narratives are really helpful to stop us being in a position where we're asserting different formulations of words, but we are not necessarily building the understanding of what those words are supposed to be working towards. And certainly in some of those interventions, felt a sense that we we're talking about the same thing, even if we're finding different ways to articulate it. In the previous session of the GGE, we raised a particular concern about the inclusion of the term "identify" in paragraph 1. And that concern is that by adding "identify" there, we don't want to be in a situation where it can be interpreted that this implies that systems are not included simply because the human commander or the human operator has themselves identified a target that they that they wish to attack and that they wish the system to engage. And I think this this comes from the fact that the term identify can be interpreted in different ways, both in terms of the human function of militarily identifying the objective that is being pursued or technologically within the sort of sequence of operational structures within the system. But I suppose what we're concerned with particularly is the very same physical system going through the very same operational process. As we would read this definition, the same system, the same operational process could fall within the characterization in situations where a commander has not identified a particular target they want to engage. but it might fall outside of that characterisation if the commander has identified a specific target that they want to engage. Not only is this risking narrowing that characterisation, but it feels like it is rather an unstable structure if the characterisation of the technological system is in some way dependent upon the mindset and the orientation of the commander who is engaging the operation. do not quite feel that 1A fixes that particular concern. I recognise, though, that the concern that I am trying to articulate is, to an extent, in tension with concerns that others have raised about not wanting to include systems where a commander has perhaps identified and marked a specific target for engagement. I think that others have mentioned homing missiles or air-to-air missile systems of a certain type that people are concerned might be captured by that. I broadly understand those concerns. Just purely for food for thought, I have a little formulation that I would like to suggest, recognising that this is just for purposes of discussion. I wondered about a phrasing as follows: the fact that human operators have set the general parameters for engagement or intended a specific target to be engaged does not exclude the system from characterisation. I think that the changes that I am suggesting there are to include a clause that recognises that the fact that a commander intends to engage a specific target should not exclude the system from characterisation. It might also be useful to shift the focus of the parameters, first of all suggesting that they're general parameters and that they're the parameters of engagement rather than the parameters of the target per se. And I wonder if that sort of shift might help to alleviate the concern that this is capturing the sort of target marking processes that are at work with certain homing guided missile systems. I also wondered if by setting the whole formulation in the past tense or the past perfect tense, It would situate this clearly as something that has happened prior to the process of identification, selection and engagement that the weapons system is engaged in, and that maybe shifting it into the past tense would solve a concern that I felt was raised by China in their intervention, as I understood it, that this could be understood as being a form of intervention within the process as described in paragraph 1. So, simply suggesting that sort of language for further consideration. And of course, I can send it to you as required. Thanks, Chair.

Thank you for that. And indeed, we would appreciate to receive it by mail for further consideration. Next on my list is Matt for peace, to be followed by the ICRC and IHL, YI. But first, Matt, you have the floor.

Thank you, Mr. Chair. Math for Peace, Development and Human Rights welcomes the current updated text as a solid basis for negotiating a treaty on autonomous weapons. Mr. Chair, there is an increasing use of artificial intelligence systems in armed conflicts, particularly decision support systems in military operations. Although they are not autonomous weapons, they raise concerns about the loss of human control over the use of force during armed conflicts. and it underscores the necessity of caution when considering future uses of autonomous weapons in armed conflicts. Thus, MATT confirms the necessity of clarifying the term "human intervention" in the text to encompass a full understanding of the mechanisms of decision-making and formulation, ensuring that human operators can review automated recommendations and verify their compliance with IHL before execution. Rather than relying solely on algorithmic outputs, the human role should not be limited to merely procedural aspects. Moreover, the text must align with the nature of software that evolves independently of human control, particularly those using deep learning. Therefore, we propose the following amendments. In paragraph 1, we suggest adding text after the term "technological components", which would read: including decision-making systems based on autonomous processing of sensor data, capable of identifying, selecting, and targeting objectives without human intervention.

Or with momentary direct control that.

Reduces the human role to mere procedural validation of machines' decisions in executing these tasks. Also in subparagraph C, we propose adding a sentence at the end of the paragraph stating, conditional that this exclusion is not used as an excuse to evade responsibility. No system should be exempt from this definition unless it is proven beyond a reasonable doubt to be under complete, effective, and auditable human oversight in each individual engagement cycle. Thank you, Mr. Chair.

I thank the representative of Mart for Peace. And next on my list is the distinguished representative of the ICRC. You have the floor.

Thank you, Chair. The ICRC would like to express appreciation to you, Chair, and to your team, and the friends of the Chair for the hard work and consultations on the rolling text. We echo the comments made by distinguished representatives from Brazil, Sri Lanka, Switzerland, Austria, Norway, and others, in considering that the text reflects many constructive contributions made by high contracting parties and observers to date. It contains the structure and substance of a set of elements that we see as a very good basis for negotiations of a legally binding instrument to regulate autonomous weapons. In the context of the working characterization in box one, We see it as necessary that the group retains a broad characterization of autonomous weapons systems that captures all systems of concern, such as the one currently reflected in the version of the rolling text. In this regard, and echoing the comments made by Austria, we reiterate our recommendation to delete the qualifier of "lethal" throughout the text, as it finds no basis in IHL. IHL protects civilians against death and injury, but it also protects objects. That said, similar to the comments made by Brazil, Italy, Mexico, and Norway, we consider the qualifier introduced in paragraph 1 to represent a constructive compromise between the proposals made throughout the past GGE sessions. Now, with regards to the critical functions of autonomous weapon systems, We welcome the progress made in building convergence and view paragraph 1a as a positive step. However, we wonder whether the current formulation fully addresses the concerns raised by delegations regarding the inclusion of identify. Indeed, we agree with the comments made by India and Japan that parameters of the targets can be ambiguous. In particular, there is a risk that it could be interpreted narrowly and not reflect other common human inputs. These might include things like setting the generalized target profiles, list of potential targets, or the operating parameters of the system. We therefore suggest three alternatives to strengthen subparagraph 1a while retaining the term "identify". First, we suggest clarifying what is meant by "identify", namely that it refers to the autonomous weapon systems matching of potential targets against a generalized target profile. which was the explanation provided by many delegations who supported the inclusion of the term "identify" in previous GGE sessions. Alternatively, and echoing comments made today by distinguished representatives of Ireland, Brazil, Panama, Sri Lanka, Switzerland, and just now by Sierra Leone, we suggest expanding the clause in paragraph 1a to list the other functions that are often carried out by humans. and that should not exclude a system from the characterisation of an autonomous weapon. This could read as the fact that a human operator sets the list of potential targets, the generalised target profile, or the operational parameters of the system does not exclude the system from this characterisation. If such an understanding cannot be reached, in line with suggestions made today by Ireland and other delegations, The ICRC would reiterate its previous recommendation to either reinsert and/or between select and identify, or alternatively, to delete the term identify altogether, as its inclusion would exclude from its scope most, if not all, of the systems that are meant to be regulated. Now, we've also listened with interest to the proposals made by Pakistan and the United States regarding paragraph 1A, and we'll provide further remarks on this later in the session. Finally, we wish to emphasize that the current text, in particular, the introduction of qualifiers in subparagraphs 1A and 1B, which accompany the reintroduction of the terms "identify" and "lethal", already reflect significant compromise made by different high contracting parties throughout the past sessions. In light of that, We emphasize that retaining those terms while removing those qualifiers risk undermining the spirit of consensus building that has guided the work of the GGE. I thank you, Chair.

I thank the distinguished representative of the Red Cross for commenting on the text, reacting to proposals and proposing alternative text to move forward. Next on my list is IHLY. I, and to be followed by Stop Killer Robots, the University of Nottingham, and ICRAC. But first, IHLYE. I, sorry.

Thank you, Mr. Chair. The International Humanitarian Law and Youth Initiative welcomes the progress reflected in the rolling text and appreciates the efforts of the Chair in engaging all stakeholders, including civil society and youth-led initiatives. We underline the importance of collective action to ensure the implementation of international humanitarian law and the principle that responsibility and accountability for the use of force must remain with human actors. Regarding box 1, the International Humanitarian Law and Youth Initiative welcomes the modifications made, including the inclusion of the term "functionally" to clarify the operational nature of lethal autonomous weapon systems, the replacement of "user" with "operator", the expansion of "death" to "loss of life", the inclusion of "identify" along with "select" and "engage", and the addition of sub-paragraph to address concerns that the inclusion of "identify" might narrow the characterisation of loss. However, based on our field monitoring of contemporary conflicts and technological applications, we would like to highlight four gaps and risks. One, potential narrowing of the scope through identify. The use of the term identify may narrow the scope of laws and exclude systems that should fall under regulatory oversight. Two, insufficient recognition of all forms of human input prior to activation. The current text does not sufficiently reflect human involvement in defining operational parameters or generalized target profiles, which is essential to ensure ongoing human control. Three, accountability gaps. The text does not clarify who bears responsibility in case of error or harm. Responsibility cannot be transferred to machines, and clear accountability mechanism for developers, operators, and states are necessary. And four, future flexibility without review mechanisms. Subparagraph C allows for future modification, but does not establish mechanism to review or update the characterisation in line with technological developments, which could create regulatory gaps. Additional suggestions to straighten the text are to: Clarify the meaning of "identify" or adjust the wording to include all forms of human input prior to activation. Reinforce continuous human control over critical system functions. Align the definition to regulatory frameworks to ensure accountability of both states and private sectors. And ensure the definition is consistent with international humanitarian law principles of distinction and proportionality to protect civilians. Thank you, Mr. Chair, for listening.

I thank you for stating the position of your NGO. Next on my list is Stop Killer Robots, to be followed by the University of Nottingham. But let me first pass the floor to the distinguished representative of Stop Killer Robots. You have the floor.

Thank you, Chair. Stop Killer Robots also wishes to thank the Chair, his team, the friends of the Chair and States for their continued efforts to build convergence on the elements of an instrument on autonomous weapons systems. We also thank States for their support for the full participation and contribution of civil society and other observers to this work. To make a brief general comment, we believe that in the rolling text, States now have a good basis with which to start negotiations on a legally binding instrument on autonomous weapons systems. We welcome that more than 40 states have already stated that the text is a sufficient basis to move forward to negotiations on an instrument and urge others to recognize this too. The agreement of legally binding rules must be the next step. Despite some important texts being lost from previous versions, which we will comment on in subsequent statements, and the fact that the text still does not directly address the critical issue of anti-personnel autonomous weapon systems and the need for their prohibition, We welcome the new version of the rolling text, which keeps key elements that reflect key components of meaningful human control. This week, we urge States to retain the progress that they have made to this point by supporting and continuing to strengthen the text to create the strongest possible basis for negotiations. Regarding box one, for Stop Killer Robots, a legal instrument on autonomous weapon systems must apply to the broad range of weapon systems that detect and apply force to a target based on sensor inputs rather than an immediate human command. The current text characterization adequately covers the scope of systems, which ranges, as states have observed today, from existing and simpler automated weapon systems to new systems in development with more complex functions. This whole scope must all be regulated by an instrument through prohibitions on systems that are fundamentally unacceptable and positive obligations to ensure meaningful human control over other systems in practice. As various delegations have observed, states must ensure that the addition of the term "identify" to the characterisation, if this is retained, does not result in an unnecessarily narrow understandings of the scopes of systems under discussion. We welcome paragraph 1A in that regard, but we also recommend strengthening it further. The text must clarify that the characterisation will apply not only to complex systems that generate targets and apply force to them based on very broad mission parameters, but also to systems where generalised target profiles or more specific targets are set by people. These systems will constitute the majority of autonomous weapon systems that are likely to be developed and used, and so must be subject to regulation. We welcome the ICRC's and other colleagues' proposals on how paragraph 1 could be strengthened and clarified. We welcome the retention of paragraph 1B for ensuring that states' understanding of the systems under discussion is not unnecessarily restricted by different interpretations of the word "lethal", a term that we continue to consider unnecessary. Finally, we wish to acknowledge that states' discussions on autonomous weapon systems do not include the standalone use of tools such as AI decision support systems that suggest targets for approval even if human input and agency may be nominal in practice in their use. However, these tools are already causing serious harm. Since they are not under discussion in this forum, states must be sure to regulate these tools legally through other forums and routes to ensure meaningful human control and prevent harm. The risks AI and autonomy in the use of force pose to civilian protection, human dignity, and meaningful accountability are grave. With more and more reports of weapons with autonomous features being used, military AI tools being linked to serious harm to civilians, and pressure growing on industry to adopt ever more lax ethical standards, despite the resistance of many technological and other experts, states work to agree on strong rules on autonomy in weapon systems and the use of force is more important and urgent than ever. We urge states at the GGE to be guided by this context as they undertake their work this week. Thank you, Chair.

I thank the distinguished representative of Stop Killer Robots. And yes, I have heard that 42 States, high contracting parties are ready to take this further. But that, for me, is also important that there are still 86 to be convinced that the text on the set of elements in front of us is agreeable. And I think our main line of effort should really be working on the text so that in the end, 128 can agree or, my interpretation of consensus, do not voice dissent with regard to what we are having on the table. Having said that, let me now pass the floor to the distinguished representative of the University of Nottingham, to be followed by ICREC. You have the floor.

My observations and recommendations are on the legal interpretation and consequences of box 1 of the current rolling text, dated 18 December 2025. In relation to paragraph 1, the term "functionally integrated" The characterization of autonomous weapon systems as functionally integrated combination of one or more weapons and technological components is meaningless and has no basis in weapons law. Firstly, it would be contrary to the obligation to conduct weapons reviews under Article 36 of Additional Protocol I to the Geneva Conventions to require either function or functionality of the weapon, as the obligation covers pre-deployment and pre-operationalisation stages of study and development of a weapon system. Remember that this obligation for the weapon review is to ensure that it complies with IHL throughout the system life cycle, and that includes design, development and use. I would therefore recommend removing the term "functionally integrated". Secondly, in paragraph 1, the reference to technological components that can identify, select and engage a target is quite surprising, because there is no reference to the emerging technologies that the GGE is concerned with, which is part of its mandate and which have been referenced in its deliberations and reports since 2017. These emerging technologies are AI and AI sub-disciplines of deep learning, machine learning, neural networks, and robotics. In fact, significant advances have been made by individual states in legislating for these emerging technologies in the civilian sector. And the GG's lack of reference to these technologies in the military sector creates an unnecessary ambiguity and a lack of clarity. It also has caused confusion about the extent of human involvement and which weapon systems are covered, as demonstrated by this morning's discussions and interventions by States. I would therefore suggest amending paragraph 1 as follows: "An autonomous weapon system can be characterised as either AI-enabled which means that it uses AI to make and implement decisions on identification, selection, and engagement of an object or person as a target without human intervention; or as AI enhanced, which means it uses AI to inform decisions on identification, selection and engagement of an object or person as a target made by a human. I am happy to send that text to you for more detail and to share with others. I think that that would address the concerns about the extent of human involvement, which weapon systems are covered and also reflect scenarios where there's a mixture, there's hybrid decision making, human machine decision making, those kinds of scenarios. Finally, just to repeat what I've said in previous sessions. Sorry, it sort of went off there. Just to repeat what I've said before, I understand states concerned about this terminology of lethality. As I've said previously, there's no requirement in IHL for a weapon to be lethal for it to be classified as a weapon. But if there is insistence on including this word, at least in paragraph 1B, reflect the terminology that's used in the convention itself and in IHL. That is, we are concerned with weapons that have effects not just to cause death, but injury, destruction and damage. Equally, I am happy to send that text to you so that it can be shared. Thank you.

Yes, the chair and his team would appreciate to receive the text in writing. Next on my list is ACREC, to be followed by Utecht University, but first ACREC.

The International Committee for Robot Arms Control would like to thank the chair and his team for the successful organization of this meeting and the continued progress on the rolling text. ICRC is an organization of academic experts and professionals and founding members of the Stop Killer Robots campaign. We would like to follow the distinguished delegates this morning from Panama, Brazil, Switzerland, Pakistan, Austria, the United States, the Philippines, and our colleagues in civil society in considering the use of the term "identify" in the characterization of laws in box 1, paragraph 1 of the rolling text, as this term was reintroduced following the discussions in the last meeting of the GGE. The word identify as specified in the text can itself be interpreted in various ways, whether intended as the common plain language version of the word, the technical interpretation of the word in the context of automated decision making, or in reference to military terminology that may vary across states and types of engagements. First, in plain language, one might interpret identify as simply a synonym for recognize. This is to say that a weapon is surveying a region, distinguishes various objects, and discerns amongst those objects that one is a valid target. This plain language definition, while easy to grasp, raises a range of technical and legal questions. Second, as a technical term, identification can have one of two different meanings. Here it can mean either that an identification of an object as belonging to a particular class in visual or other data, i.e., the identification of a sighted object as belonging to a type of a pre-programmed sort, an identifiable group of people or things. Or, alternatively, it can mean the recognition of a specific individual within such a class, i.e., the recognition of an individual person, perhaps by the way of biometric data. This makes identification somewhat unclear in terms of identification of targets and how that relates to specific individuals or whole classes of potential targets, raising the question of which of these might be beneficial to include in a criterion for defining laws. While identification can be seen as distinct from selection, it does not necessarily add clarity to the characterization of laws. Third, we can look to military doctrines for the definition of identification within targeting processes. For example, positive identification in the NATO joint targeting doctrine includes both observation and analysis, specifically including legal considerations. The identification of a target as a valid military objective, PID, as it's called in this terminology, must be upheld while a target is under engagement or damage is being accessed. This use is most closely related to international law, where identification is a condition for IHL compliance, such as the distinction requires identification of lawful targets as a prerequisite to the use of force in armed conflict. An unfortunate implication of including such a characterization of laws might be that such systems are somehow seen as being capable of making such legal determinations, or that they would only select and engage targets that are legal by definition. ICRAC is therefore of the opinion that the possible inclusion of a function such as identification that can be seen to imply legal analysis or evaluation as a criteria for defining laws would be improper so far as automated systems cannot make legal assessments or judgments and should not be considered as making legal determinations. Further, including this as a required characteristic of laws may render the potential instrument inapplicable to most of the systems currently being considered and developed. While they support the suggestion of the distinguished delegates this morning from Italy and Switzerland that as states engage in negotiations for a legally binding instrument, they should further clarify what is meant by this term in this section of the text and how it differs from selection alone, or that we should return to a more straightforward characterization of select and engagement. Thank you, Chair.

Let me thank the distinguished representative of ICRAC and now pass the floor to the distinguished representative of Utrecht University. You have the floor.

Thank you, Chair. I'm delivering this intervention on behalf of the delegation of the Asia Pacific Institute for Law and Security and my own delegation, Utrecht University. First, we would like to thank you and your team for your tireless work in getting the text where it is today. We also thank you for your engagement with civil society and academia during intercessional period. Chair, we support the thrust of box one on the characterization of laws. Almost every word in the characterization reflects several years of painstaking discussions. It captures the meaning of laws with considerable precision, yet has the virtue of remaining concise. Subparagraphs A and B helpfully clarify some elements of the characterization that several delegations have found critical. Those sub paragraphs wisely abstain from abstract definitions of notions such as identify and lethality, but focus on the scope of characterization. Some delegations raised questions about the phrase functionally integrated, which we have previously supported. The challenge here lies in ensuring that the characterization of laws captures weapon systems rather than simply weapons. The word functionally simply ensures that the combination of one or more weapons and technological components does not need to be integrated physically into a single device, but must be integrated into a single weapon system in terms of the functions of individual technological components. It is this integration of functionality that allows the weapon system to identify, select, and engage a target without intervention by a human operator. With respect to subparagraph A, we support the suggestion by the ICRC to replace the reference to the parameters of the target with a broader reference to the list of potential targets, the generalized target profile, or the operational parameters of the system. An alternative and more comprehensive formulation we would propose for this paragraph would be the fact that a human operator sets the parameters that comprise the profile of the targets does not exclude the system from the characterization. Chair, we continue to have concerns about the phrase within the scope of application of the CCW included in the square brackets. For one, it's unclear why it's necessary, considering that the GGE operates under the auspices of the CCW and within the mandate given to it by the Meeting of High Contracting Parties. But perhaps more importantly, the expression "scope of application" has a specific technical meaning under Article 1 of the CCW. This article defines the scope of application of the Convention as the circumstances in which the Convention applies as a matter of law, namely situations of international armed conflict, and following the amendment to Article 1, non-international armed conflict. Using this phrase in the present context, therefore, imports a treaty concept tied to the issue of the application of the convention. Put differently, the characterization of laws risks becoming enmeshed with the question of legal characterization of armed conflicts. This is not how the CCW and its protocols have traditionally regulated weapons. The convention regulates weapons based on their characteristics. That is, their potential to cause superfluous injury or unnecessary suffering or of indiscriminate effects. The current wording, however, appears to unnecessarily link the characterization of the weapon to the circumstances of its use rather than its inherent features. We suggest, therefore, that the phrase "within the scope of the application of the CCW" be deleted. Furthermore, the phrase for the purposes of these elements of an instrument is perfectly sufficient to limit the application of the characterization to the present document. This more concise language would also mirror the simple phrase for the purpose of this protocol used to introduce definitions in Article 2 of Protocol 2 and Article 2 of Protocol 5. We also consider that the phrase can be characterized as creates some ambiguity. The word "can" suggests that an AWS could be characterized in the way contemplated in the rolling text, but that it could also be characterized in some other way, possibly in a manner that has not been considered by the GGE. We submit that the simple word "is" would be sufficient to replace the phrase "can be characterized as." Thus, the beginning of the paragraph could read, "For the purposes of these elements of an instrument, a lethal autonomous weapons system system is a functionally integrated combination and so on. This would not prevent, of course, the GGE from amending the characterization later. I will submit these comments directly to you and your team after this intervention. I thank you, Chair.

Let me thank the distinguished representative of Utrecht University and also for sharing the suggestions she has made. This was the end of the, no, I see, indeed, I was just on the point of saying that, of course, The floor is open for a second round also to react to previous statements. And I recognize Austria, and I see Pakistan and Japan. And that's it for the moment, I think. But let me first pass the floor to the distinguished representative of Austria. You have the floor, sir.

Thank you for giving us the opportunity to have a second round in a good old tradition that we have in the TTE laws that we also react to each other's proposals and also to the proposals and suggestions by the civil society, which I have to say in this round, they're really rich and much to digest. So there have been some valid points that we also have to take into account and we move forward with the text. For now, I will focus more on the issues that have been brought up by other delegations and also maybe one or two general points that have already also been made by Norway and Pakistan about the purpose of this characterization. What we have in front of us is a working characterization which we are using for a specific purpose, which means to develop a set of elements of an instrument, which can in the long run be something else. But as we have consistently pointed out, is that this is a fluid document. When we move forward with negotiations or with other changes in the following text, this also has to be reflected in the characterization section because it's interlinked. This is also related to the point that in general, what we are setting out here has to be broader than what comes later in the text. This is not the prohibition and the regulation section, this is a definition characterization section. And the part where we actually decide this system is a problem, this system needs to be regulated, is later in the text. And this just has to be reminded because I think we are falling into this trap time and time again. Another general point I also have to make again is that what we are trying to do here is to cast a wide net and tighten it later. because it's much easier to have a broad characterization that is then specified and pinpointed down later on than have a very narrow characterization that is full of loopholes. And some delegations who are here pointing to specific weapon systems that are here in this net and should be out of this net, they are plugging little holes into this net as far as we see it, or to some extent they are doing more spearfishing than actually net fishing as far as we can see. Not to react to some proposals that we heard, I think we had very long discussion and different versions on para one. We would not say that necessarily the proposals that we heard are bad or are not helpful, but going into a new conceptualization of this whole para could create a lot of problems and we would rather prefer working on this version or keep it as it is. instead of doing large rephrasing as, for example, proposed by the US delegation. Another point that is important to us and that an argument that was made, it's also not new, but it is related to the issue of designed to. It is true that we have three protocols that are using this formulation as well, but it's also always important to look into the specific context of these protocols. I mean, protocol IV, for example, talks about, is in general about a specific function of a weapon system, blinding laser weapons, so it's very clear that the design to make sense there. When it comes to Amendment Protocol II, we are talking here about definitions that are taking a certain existing system, munitions or devices, and specifies with their design how they function, how they then turn into anti-personal mines or other devices. So we are talking here also about a very functional approach. The problem that we are facing here is that we are building a fundamental definition of a system and not building up on something else where we are adding a specific function of a weapon system. So this is to some extent different. The third point, and this brings me to Protocol 3 on incendiary weapons, there is a major loophole in Protocol 3 which has caused a lot of discussions and also problems. And there is a high risk that specifically in this field we are also creating a new loophole by saying only weapons that are designed for this and this purpose will be under this characterization. Even more so as we are talking about a segment or field which is characterized by dual use like no other has been so far. This is a huge problem and sorry for being so long about this. Then already on point A, this subpara was introduced just recently so I think it's obvious that we still need to work on this and that it is still fluid. I mean, there have been several arguments how to improve. and better characterize it. We are happy to look into all these proposals. The general problem that we are trying to solve here, and this also has to be pointed out, is one that originally did not exist. So the identified problem has been brought into this discussion because there has been some issues and also some problems which said that this is an operational significant element and is also necessary to include some kind of weapon systems or to exclude some others. As we have pointed out in the past, identify for us is included in select, but we are open to resolve this issue via A. But this also means that we have to resolve the underlying problem, which is that the word identify has two different meanings, and as we just heard, there could also be a third meaning, what you hear from Article 36 and others, or ICRAC also made some valid points in that direction. So we have to resolve this and be very clear in what way we mean, what kind of meaning we attribute to identify, and that is the purpose of A, which means we have to be thorough, we have to be careful. and also make sure that we are not also here creating a loophole. And the issue that has been also discussed for many years now is the issue of nominal human input, also the possibility that just via clicking on a certain button or something else, create validation of decisions and a validation of an autonomous weapon systems, which in this case could also exclude it from a certain characterization or from a whole instrument. So this is an issue and we are happy to look into the different ways how to resolve this. A reaction to the Russian Federation, which said what we are doing here might not necessarily be a set of elements for an instrument. For us, it has been very clear over these three years that this part is the set of elements of an instrument part and not the possible other instruments part. We have not seen too many working papers that set out other possible measures. So everything that we have in front of us in all the working papers are pointing towards a set of elements of an instrument and that's quite clear for us. Last point, sorry for being so long, there have been some concerns and we definitely shared a concern that was also brought up by Israel that we here create a new definition of lethality, which does not exist in IHL so far and many colleagues from the civil society and from the ICRC have pointed out why, and that we are kind of touching new ground here, which we probably should not do. But as pointed out before, this is a compromise. This is a way to resolve differences that we all had for many years, and we are eager to look into this, and the current text is probably the best solution in doing this. But one important point which has been alluded to today as well is that some colleagues have mentioned that we always had legality as a core element of our discussion. As someone who has been here for five years, I can, uh, point, I can, uh, I can assess, uh, and also confirm that this has always been disputed. that the issue of lethality has always been an issue and that it has been unclear how it is meant. And in a certain phase that we have gone through, we also had the situation that only a handful of delegations still thought that the lethality element is relevant and all others did not deem it as important as it suddenly is again, uh, now. So the, the idea that this has been a core element since the beginning of the GGE laws is factually, uh, wrong, and I can testify in court and anywhere, um, that this is the case. Um, we also never had, uh, a discussion about if we are talking about anti-material, anti-personnel laws. I mean, the idea that we would exclude anti-material is completely new, and I hear about it here, I mean, today, the first time, at least the word "manpower" is completely new to me. These are some points we just wanted to make to bring the discussion forward or make it more complicated. But in any case, thank you for your attention.

Let me thank the distinguished representative of Austria for sharing some further views with regard to the working characterization and also his view on the proposals made and especially the issue of identify. Next on my list is Pakistan, to be followed by Japan and Colombia. the Russian Federation. But let me first pass the floor to the distinguished representative of Pakistan. Sir, you have the floor.

Thank you, Chair. My delegation also welcomes this opportunity, and we think that by providing this opportunity to react to earlier observations, we are trying to build convergences and find middle ground. So we really appreciate that. We have listened very carefully to the observations that have been made by several colleagues this morning. And I would like to limit my intervention only on the issue of paragraph 1A. So from our perspective, we have listened and we have heard four distinct concerns. Let me address each one by one. So first, we have listened to this argument that explaining identify requires us to also define, select, and engage. Chair, paragraph 1A, how we see it, is not a definition of identify. It does not tell us what identification means and at which stage it occurs or how a system performs it. It does something much narrower and more targeted. It closes a specific interpretive risk that arose in this room, the risk that a state could claim its system falls outside the characterization. simply because a human operator preset targeting criteria. That risk was unique to identify in our view. We have, and colleagues will recall, that we have heard broad convergence on select and engage. We have not heard any ambiguity on these terms. Therefore, we think that where the shoe pinches is where we apply the remedy. We do not need to bandage a limb that is not required, that is not injured. The logic that we cannot clarify one term without simultaneously clarifying all terms, I think it would be a recipe for paralysis. Second, Chair, we have also heard a suggestion to recast paragraph 1A in positive light instead of its double negative connotation that exists. We understand this instinct, and I think we also appreciate the constructive spirit of this suggestion. However, the negative formulation is preferable precisely because of this restraint, if I may use the word. By stating that setting parameters does not exclude a system from the characterization, we are aiming to close a specific loophole without prejudging broader questions. We are not defining what identification is. We are not prescribing at which stage of the targeting cycle it occurs. We are not telling states how their systems must work. We are saying you cannot use the fact of human parameter setting as an escape clause. A positive formulation, we think, would require us to say what identification is, which is a definitional exercise that we have spent for several years in this GGE. The moment we start writing identify means X or identify means Y, we lock ourselves into language issues. And also, notwithstanding the pace of technological change that will continue to happen. So we think that this formulation is technology proof. Chair, the third concerns that we have heard on 1A is that it could inadvertently capture some systems that exist, and we have listened to examples of fire and forget homing munitions or GPS guided missiles? We think it cannot, and I'll explain. And it requires a reading of paragraph 1 and 1A as a unit. So how we see it, paragraph 1 sets the threshold. A system that must be able to identify, select, and engage a target without human intervention in execution of those tasks. Now, a conventional fire and forget munition, or a GPS guided missile, where a human locks onto specific vehicle or launches a GPS guided missile to certain human designated coordinates does not meet this threshold. The human has already done the identifying and selecting function. The munition will perform guidance and engagement, not autonomous selection among competing targets. Therefore, because such systems do not satisfy paragraph one, paragraph 1A simply never applies to them. 1A is not a net that pulls systems into the characterization. It prevents systems that do meet paragraph one from being pulled out by invoking upstream human involvement. So therefore, we think the sequence matters in paragraph one and 1A clause. Chair, lastly, fourth, we heard the suggestion that perhaps targeting parameters or parameters itself, this is a broad term and it's ambiguous. We do have sympathy with this argument. We take this concern seriously. And in fact, it points towards strengthening 1A rather than weakening or deleting it. In practice, how we see it, We understand that human operators interact with weapon systems at varying different levels of specificity. For instance, at one end, a human designates a specific target, precise coordinates, for instance, a particular vehicle. On the other end of the spectrum, a human could define a target category. For instance, engage any radar emitter matching frequency X with grid square Y. And the system autonomously determines which specific object to strike when and where. So when a human fixes a specific target, there is no autonomy in the selection function, and the system already, as I said, does not meet paragraph one. We agree with this entirely. But when a human sets general criteria or general parameters or a target profile, a class of objects or a geographic zone, this is where the machine makes the specific targeting decision. And this is precisely the autonomous function paragraph one is designed to capture. And Chair, that's why in our first intervention we proposed that perhaps in addition to using parameters, we could add the word or profiles. to bring further clarity and strengthen this language. We have also heard very closely, I think, the suggestion made by Utrecht University that was also going in the similar direction, and we support that. And we think that this small addition costs nothing in terms of narrowing, but it closes the door more firmly against the argument that a preloaded target profile is qualitatively different from parameter and therefore not covered under 1A. So in conclusion, Chair, I think we are very close on the issue of paragraph 1 as drafted. It does exactly what it should. It protects conventional guided munitions, not by exempting them, but because they never meet the paragraph 1 test in the first place. And it ensures, most importantly, that genuinely autonomous systems cannot escape the characterization through a formalistic argument about the human involvement and where it occurs. And lastly, Chair, before I finish, I would just like to echo that on the other suggestions that we have heard regarding further amendments in paragraph one, we would like to echo the views of the delegate of Austria. We think that we are very close in terms of finding convergence on this issue and introducing any new significant changes or proposals at this stage would only take us further away from that convergence. I thank you once again.

I thank the distinguished representative of Pakistan for responding to remarks made earlier by other delegations. Next on my list is Japan, to be followed by Colombia. But let me first pass the floor to the distinguished representative of Japan. Sir, you have the floor.

Thank you, Mr. Chairperson. My delegation carefully listen the argument this morning session and this afternoon as well. So let me take the floor again in order to add just a couple of points quite briefly. Point number one, regarding identification, that is essential characteristics about loads. As we have stated in the previous and various occasion in the meeting, identification is a function that directly connect to the determination of whether something is a military objective. And we consider it important for implementing the principle of distinction under IHL. So within the targeting cycle and discussion on loads, identification refers to the process of searching for and determining from memory database, et cetera, and what matches or what has been captured by visual observation, sensors, electric signals, and we consider it to be the function that distinguishes between civilian or civilian objects and military objectives. And Mr. Chair, point number two, quite briefly, regarding lethality. While my delegation does not object to including non-fatal injury and material damage within the scope, but we would like to clarify the focus on this discussion on the lethality as included in the very mandate of this GGE. And finally, regarding this paragraph 1A, In this morning's session, my delegation just mentioned that the formulation of a human operator sets the parameters of the targets. It's a bit ambiguous. So in this regard, my delegation is of the view that U.S. proposal on paragraph 1A is worth considering and would like to study the proposed languages by the United States. Thank you very much, Mr. Chair.

I thank the distinguished representative of Japan for reacting to proposals made by other delegations. Next on my list is Colombia, to be followed by the Russian Federation. But let me first pass the floor to the distinguished representative of Colombia. Madam, you have the floor.

Gracias.

Thank you, Chair. This is the first time that we're taking the floor, so we'd like to thank you and your team and the secretariat for all the efforts made to produce this rolling text that we're discussing this week. Colombia considers that the rolling text is a sufficient basis upon which to begin our next stage in this process, which we consider should be the negotiation of a legally binding instrument. And therefore, we have joined the call of other countries to that purpose. We listen very carefully. Apparently, we need another 80 countries to... We need to convince another 80 countries for this, and we consider this week to be absolutely crucial. Our comments are also going to be very specific this week. We're going to focus on some points that we think still need adjustments in the rolling text, and our silence will be synonymous with agreement with what is proposed. However, as other delegations have said, this week's exercise should not lead to what you have proposed in the rolling text being further diluted. With regard to box number one and the text therein, We believe that deleting the qualifier "lethal" in the rolling text could be an option. However, the proposal that you have put in paragraph 1b with the inclusion of "loss of life", as other delegations have said, would be another option for addressing this divergence. Other delegations have mentioned their views about the inclusion of the words "identify" or "identification" in paragraph 1, and we have also heard alternatives to the inclusion of those words. We are happy to work on various proposals, on the understanding that we cannot limit the scope of this rolling text. We also listened very carefully to the various contributions made by civil society representatives and academia. We found their contributions very interesting, but I will mention just one, which was made by the University of Utrecht, about part of what appears in square brackets in the first part of paragraph 1. and which relates to the words within the scope of the application of the CCW, we consider that they made a very valid point that we ought to take into account, especially in order to ensure alignment and consistency between what is being discussed in the rolling text and the Convention per se and other conventions. Thank you.

I thank the distinguished representative of Colombia for sharing the positions of her delegation. Next on my list is the Russian Federation, and that is the last one on my list, so if there are any other delegations wishing to take the floor, please indicate so. But let me first pass the floor to the distinguished representative of the Russian Federation. You have the floor.

Sir.

Thank you, Chair. We had not planned to take the floor in the context of the interactive dialogue again, however, following the final part of the comments made by Austria, we felt compelled to take the floor. We categorically disagree with the assertion that the criteria of lethality, the characteristic of lethality, was not at the heart of the discussions of this group throughout its existence. That is absolutely not the case, and it is a misleading statement to make, and it creates an erroneous understanding of the situation in connection with what this group was created. The characteristic of lethality was a fundamental and foundational concept at the stage of the formation of this group, even prior to its actual establishment. And the characteristic of lethality is specifically the characteristic which, on the basis of which, the discussion was begun regarding the need to discuss this topic within the context of the CCW specifically? It was by positioning the risks of striking human targets with regard to this category of weapons, that this discussion was initiated on new technologies in the area of lethal autonomous weapons systems within the forum of this Conventional and Certain Conventional Weapons. That is why this term appeared and was largely a compromise which made it possible to create this term in 2016. Furthermore, in the context of the group itself, the discussions of lethality did not, during the discussions, it did not, it was not questioned until 2021, when for the first time this question was raised. Prior to that, nobody raised any questions regarding the fact that the characteristic of lethality, which is included in the term lethal autonomous weapons systems, should be one of the fundamental characteristics of this category of weapons. This term was not diluted. There were no attempts to try and say that we're not that we should talk about non-lethal autonomous weapon systems, but simply about lethal autonomous weapon systems. Therefore, I apologise for setting the record straight in that way, but we believe that it is essential to make sure that things are understood correctly so as not to mislead other colleagues. regarding the reasons for which this group was created and what happened in this group up to a certain date. On that basis, our delegation, like other delegations, considers that the characteristic of lethality remains one of the fundamental characteristics in the context of the discussions of this topic within this group of governmental experts. Lethality absolutely must be included and explained in the definitions of these weapons systems and attempts, which is made in the context of paragraph 1. Thank you.

I thank the distinguished representative of the Russian Federation. for stating the position of the delegation responding to remarks made earlier by another delegation. Next on my list is Costa Rica. Let me pass the floor to the distinguished representative of Costa Rica.

Thank you, Chair. Costa Rica would like to commend your leadership at this stage, Chairperson, at the helm of the work of the Group of Government Experts under this Convention. And we are grateful for the rolling text, which in our view is a good basis for moving forward on negotiations on a legally binding instrument. This is the final year of the mandate given to this group, and that is a critical opportunity for us to show that multilateralism is a good thing. to realism can respond effectively to the challenges posed by emerging technologies in the area of lethal autonomous weapon systems. For Costa Rica, whose foreign policy is based on the promotion of peace, disarmament, and the strengthening of international law and the protection of human rights, This debate is not merely technical, it is deeply ethical and legal in nature. Regarding box one, we welcome the functional characterisation of the capacity, which is based on the capability of identifying, selecting and engaging a target without intervention by a human operator in the execution of these tasks. We feel that it focuses on the critical functionalities and offers a good basis for starting. Regarding paragraph a, we welcome the fact that it clarifies the fact that a human operator sets the parameters of target and that this does not exclude the system from the characterisation. It is important because pre-programming does not amount to human control necessarily when actually executing the task, the strike task. We also believe that lethal is important, the idea that these weapon systems can cause harm to an object without necessarily creating, causing death. would be a preventive measure in keeping with international humanitarian law. Costa Rica believes that this box is an essential pillar for the text and it is a clear, functional and legally robust characterization which is essential for ensuring that the future of this instrument meets adequately the challenges that these technologies raise for the protection of civilians, for human dignity, and for legal responsibility and accountability. Thank you.

I thank the distinguished representative of Costa Rica for stating the position of his delegation. Please, may I remind delegations not to speak too fast, because this was about, I think, what the interpreter could do at most. And I think we all need their help to understand what is being said in the room. Next on my list is Brazil. Let me pass the floor to the distinguished representative of Brazil.

Chair, I would like to react to some of the comments and proposals made this morning and early this afternoon. And at first to thank many of the constructive proposals that we heard. As for the amendments that we heard on advanced for paragraph one, like others, we would strongly caution against any major changes, especially ones that would have an impact on the scope of systems to be covered under this characterization and that would introduce novel elements here. On the amendments proposed by the delegation of the United States to sub-paragraph 1, we'd like to note our support to the intervention by Pakistan regarding our preference for the formulation that we currently have based on a negative statement, which is very eloquently explained by Pakistan, offers advantages from the point of view of clarity and including also with regard to the issue of fire-and-forget munitions. However, we will study the US proposal very carefully. And we would like to note that we welcome the approach followed by the US in listing systems that could be in the gray zone of this characterization due to the inclusion of the identify function. And by stating clearly that it believes that these systems, such as active protection systems, close-in weapon systems, lighter munitions and air defense systems should be part of this characterization. We urge all delegations that wish to amend this text to follow the same transparent approach and state very clearly what types of systems they wish to exclude from this characterization. That would allow us to engage more constructively in this discussion. Regarding lethality, we would like to support also the comments made by Austria. It has been understanding that the vast majority of delegations have been expressing in the past three years at least that the inclusion of this word might lead to confusion. And I think that the discussions we've had since have illustrated potently the effects of this inclusion. While it is undeniable that this word has been in the mandate of the GGE since the very start, it is also undeniable that we never had a common and shared understanding of what this means. So it is important that we pursue, again, a transparent approach and state clearly what we understand by this term and find a formulation that would allow all of us to be comfortable. I thank you, Mr. Chair.

I thank the distinguished representative of Brazil for also reacting to earlier remarks and stating clearly the position of his delegation. That's what I have so far with regard to requests for the floor on box 1. So let me just give a last opportunity. If not, then I think we This ends our discussion on box 1. I will not try to summarize what is then usually called a very rich discussion, because I think it was a rich discussion. And I think also that what I said already in my opening remarks, a few issues clearly deserve further attention. whether it's still divergence in the room. The issues highlighted, I think, already during the informal consultations, that indeed, with regard to what is now in the text on the critical functions and with regard to where we included again, identify that we would need some explanation. We had a discussion on whether this explanation is necessary because that could trigger further definitions. I think that that has been discussed very, very extensively. And I think for a large part of the room, having identified in the text should be accompanied by something, and that something I tried to grasp in 1-alpha. That text is certainly not holy to me. But I get the impression that some kind of an explanation is holy to quite a large part of the room. So there we have to find each other. And there's clearly, um, that is clearly some, some more work to be done, being done in the coming days. And I think also with regard to, um, the issue of lethality, um, quite a number of delegations in the room, insist on it, and we have heard explanations for that. I'm not going into the detail. I think it's clear that for another part of the room, they can only agree to it if there is something, and that something is one bravo. And again, one bravo could read differently in the end. The chair is pretty agnostic to that as long as you all agree on it. And I think it was eloquently phrased also by one delegation. I think we would make a contribution to the wider discussion if we, in the end, would then have something how to deal with lethal. With regard to 1-Charlie, This has been said in the past. We should be careful how much time we spend on it here. Of course, the text with regard to 1C could be refined, but in the end, this is something for future consideration. And with regard to future consideration, let me just repeat what has been said already a few times today. Let us not look too far ahead. Let us all stay focused. on what we have to deliver by the, what is it, the 4th of September later this year and not speculate on what comes afterwards. And then I'm now going to get my speaking notes with regard to box number two, if you please bear with me for a moment. Thank you for that. Box number 2. This box enjoyed broad support during our last formal session, so it was my mission also to retain as much as possible while making some refinements to further improve the text. One of the main changes is that the notion of context-appropriate human judgment and control, previously contained in paragraph 5, has been moved to box 3. This is not to diminish its importance, as it remains an essential pillar of our work. However, this box addresses existing box number 2, addresses existing IHL obligations, while the notion of context-appropriate human judgment and control is a law-specific notion that, as the text says, is needed to ensure compliance with international law, in particular IHL. Further to this, I wish to reiterate that the boxes themselves, as you see them in the rolling text at the present, will be removed later on. As a result, it does not matter in what box a certain formulation sits, but rather where in the greater flow of the text it makes most sense. In the case of context-appropriate human judgment and control, we considered the better placement to be slightly further down in the text. We will be able to discuss this change in further detail later when we address box number 3. So I kindly ask you to focus on the remainder of the paras for now. On para 3, it reiterates that IHL imposes obligations on states, parties to armed conflict and individuals, and not on machines. It's not new language. It was originally in box number 5. However, as requested by many of you, I have decided to move it up and to integrate it into box 2, as it is more relevant to the application of IHL than to the other aspects covered in box number 5. Besides these changes, Reference to core IHL principles, such as limits on means and methods of warfare, temporal applicability and Mortons Clause, remain unchanged. I believe that, thanks to these changes, Box 2 is more legally sound and foundational, and it anchors the rest of the text in established IHL principles. I would now like to invite delegations to take the floor and to provide observations and suggestions. The floor is open. Should I take from the room that there is no appetite for... Ah, I see. I refrain from stating the rest of my sentence. I see the United States And I saw a request from Stop Killer Robots, I think. But let me first pass the floor to the distinguished representative of the United States. Sir, you have the floor.

Thank you, Chair, for giving us the floor. And I'm sorry we hesitated. We were just-- we were discussing your introduction, which did I guess, create some questions in our own mind. I know we've had a lot of conceptual discussion about what we're doing with the boxes and what are the general purposes of the boxes. And your clarification just now that, you know, eventually the boxes are going to be going away. And so, you know, for you at this point, this is mostly just about the flow of the provisions. doesn't necessarily match the way we've been thinking about this. So it's very helpful to have your clarification on this point, and we will certainly further discuss it within the delegation. And I will just say that the way that we have framed the intervention on this box is generally following the way we have understood the logic of the use of boxes in the rolling text. So what you will hear will not necessarily correspond with what you've just presented as your vision for the sequence of the document. I think for us, we have largely seen Box 2 as a place to identify more general, and I think from our perspective, what we would expect to see eventually in more of a preambular sort of portion of any future instrument that are laying out general ideas And we really had viewed box three as the place where, as you just put it, the sort of fundamental IHL principles are laid out. I mean, for us, that's where we see them. That's where we see the cardinal principles of IHL. That's where we see the reflection of the prohibition against superfluous injury and unnecessary suffering, et cetera. So I think there's a lot to sort of think about and reflect on, given the way that you've just framed the flow of these boxes. So having said that, just to walk through some of the points that we had wanted to present in this afternoon's intervention. So as I've just said, I think we had viewed box two as a place to reflect some of the foundational understandings that are critical in understanding the remaining boxes, three, four, and five. and the prohibitions and the other measures that are pretty clearly identified in those boxes. And because we continue to have, Chair, just a fundamental concern with the approach on context-appropriate human judgment and control that is reflected in those boxes, boxes three, four, and five, we did want to come back to the idea of positioning The concept that we think people are trying to achieve in referring to context appropriate human judgment and control, we thought that those are ideas that would be better addressed in box two. So first, as we've said consistently, and I think we continue to hear this in interventions from the floor, there really doesn't seem to be a common understanding on what context appropriate human judgment and control means. whether it's being proposed as a new or as an additional legal requirement. We've heard some delegations even this morning in their interventions on the characterization box refer to these five words as really just a restatement of meaningful human control. And I think that is a very clear signal that for those delegations, when they see this concept reflected in this document, no matter where it is, That will essentially be a way to bring into the document and some interpretive matter what those delegations have been saying meaningful human control is and represents. We have disagreed with that consistently. For us, distinction, proportionality, and precautions in attack are the operative rules, not context appropriate judgment and human judgment and control. And to react to a framing that we've also heard today that this is the tool for ensuring compliance with IHL, I think we disagree with that as well. It's a tool. It's -- we understand that it couldn't be -- and we've said this pretty consistently over the years, that this is an idea that can be one way to get to compliance with IHL, but it's not the exclusive, it's not the indispensable principle, and it's certainly not a legal principle. At the same time, we do understand the desire to capture the importance of control with respect to lethal autonomous weapons systems and to include language that elaborates on what is needed to comply with IHL. And as we have said in the past, and as was reflected in the proposals we made last year, we think such language could go in box two, with other, again, our perspective, preambular concepts, we don't support making this language a chapeau to a list of measures such as currently exists in Box 3. So, given that general concern and with the aim of providing additional framing for the content in Boxes 3 through 5, we think there are a number of paragraphs that we could consider including in Box 2. First, we think we should consider a paragraph that reflects a core agreement from prior GGE reports where we have said in the past, and here I'm quoting, "emphasizing that control with regard to laws is needed to uphold compliance with rules of international law that govern attacks." This language describing control with regard to laws was a carefully negotiated formulation. And we think we should capture that in this document. Second, we could consider a paragraph that combines notions from box two, paragraph four and box three, paragraph six. It could read further emphasizing that measures, including during development, relating to control and human judgment are needed to ensure the use of laws is in compliance with the principles and requirements of distinction, proportionality and precautions in attack. This language recognizes that autonomous weapons are different from other weapon systems because the system software that enables autonomous functions is created during development. As a result, we think it is appropriate to highlight that measures during development are needed to ensure compliance during use. A third proposed addition is to include language from based on guiding principle D, which emphasizes accountability and the importance of human command and control chain that language could read. confirming that accountability for developing, deploying, and using any laws must be ensured in accordance with applicable international law, including through the operations of such systems within a responsible chain of human command and control. And finally, Chair, just on this issue, as High Representative Nakamitsu noted in her remarks, we are in a time of rapid technological developments. We strongly believe that this document must not stigmatize technology, To provide balance to the document, it should recognize that technological advancements can be beneficial, for example, by improving the protection of civilians. We propose drawing from guiding principle H, which the GGE has developed and was endorsed by the CCW high contracting parties. Our proposal is recalling that consideration should be given to the use of emerging technologies in the area of laws and upholding compliance with IHL and other applicable international obligations. Now, Chair, in addition to the proposals I just described, we do also have some thoughts we would like to share relating to the language that's reflected in this text. We appreciate that these paragraphs lay out foundational principles of international law and serve a preambular purpose. As I say, from our delegation's perspective, we had been viewing this as a place to anchor the rest of the document through preambular language. And so we would suggest that we could frame these paragraphs with preambulars, such as we're calling or recognizing. And so I know that's not your vision, but to the extent that there's room for discussion of this issue, I think that would be our delegation preference. Additionally, for paragraph one, we would propose adding to the end, and that IHL applies fully to the development and use of laws. Moving this language up from paragraph four will contextualize paragraphs one, paragraph one's recollection that there are limits on a party's choice of the methods and means of warfare. In paragraph two, we have some legal concerns, as we've explained in previous meetings. The principle that we think is most pertinent to our work on laws is that the rules on planning and conducting attacks apply regardless of the technological means used to conduct the attack. This principle has been reflected in past IHL treaties, such as the Hague Conventions. We recommend highlighting this principle in box two because it explains why the use of laws must comply with the existing IHL rules governing attacks, even though these rules were formulated before the development of laws. We also welcome the addition of paragraph three, which was pulled up from language that was in box five. We believe other paragraphs might benefit from being placed in this box, such as paragraph one of box five, because they contain preambular ideas. And finally, Chair, and this is our last point. Oh, sorry, not our last point, close to our last point. We would recommend adding to the end of paragraph three, the following language. and that the rules of international law that govern attacks require, inter alia, that human beings make certain judgments in good faith-based on their assessment of the information available to them at the time, and that they exercise due diligence in the implementation of these rules. This would draw from previously agreed conclusion. It highlights the helpful concepts of good faith and due diligence, which are critical and well-established requirements in the implementation of international law It also links those concepts to the facts that obligations are tied to states and individuals. Chair, I know that this was a lot. I apologize for the length of the intervention. We will circulate these suggestions to you and others. However, it is important for our delegation to make clear both our serious concerns about the current language and our willingness to work creatively and collaboratively with the other delegations here to find some common ground towards our goal. of a successful and substantive report in September. Thank you.

Let me thank the distinguished representative of the United States for sharing the views of his delegation with regard, not only with regard to box 2, because he went way beyond that. I think you went indeed way beyond that. Yes, I'm tempted to quote the mandate we have, of course, and that is about a set of elements, not about a set of elements of an instrument, because if we now start also using preambular language, then it looks like we are starting to draft an instrument, and so far I have resisted against that, because that would clearly trigger some reactions from other side of the room. and that would distract from the work that we are doing. We have had earlier discussions with regard to the boxes and with regard to moving text from one box to another. I have to be careful with that, because we have also been moving text around, but I think that I gave an explanation for that. We So far, the intention is to get rid of the boxes, but have a logical flow. So that would still matter then where text is in, where a certain element is within the set. So that having said, and indeed, we would look also forward to receiving your proposals. There's quite a lot of new text. I see what is quite interesting, of course, that a number of the issues you proposed refer to the issue of control, refer to the issue of human judgment. We put that in one element. the one on context appropriate human judgment and control that is presently in, what is it, para 6 of box number 3. That is a different approach, of course, but we will carefully study your proposals and look forward to receiving them. Now, looking at my list, I have Three more requests are also for from Stop Killer Robots, Sweden, and Pakistan. Let me first pass the floor to Stop Killer Robots. You have the floor.

Thank you, Chair. Regarding box 2, Stop Killer Robots welcomes the addition of paragraph 3 to this box, which, as you noted in your introduction, has moved up from box 5 in the last version of the text. This paragraph gives important recognition to the fact that legal obligations are addressed to and must be fulfilled by people.

And that this cannot be done by machines.

This is of central importance. This paragraph should inform how other elements in the rolling text are read and understood, and the future elaboration of legal rules based on this text. Stop Killer Robots also welcomes, in paragraph 5, the continued retention of the Martens Clause in the text. Noting the US delegation's comments, we will be giving comments on context-appropriate human judgment and control in our intervention regarding box 3. I thank you, Chair.

I thank the distinguished representative of Stop Killer Robots. Next on my list is Sweden, to be followed by Pakistan. But first, let me pass the floor to the distinguished representative of Sweden. You have the floor.

To take the floor, since we can express support for para 1 to 5 in the box 2. But along the same lines of the argument of the US, from Swedish side, we do miss the element of context appropriate human judgment and control, which were highlighted the role of the human in relation to laws and served to anchor this notion in IHL and the rolling text. and we believe that it was a proper place to do this in box 2. But if moving the para to box 3 brings us closer to reaching consensus for the text as a whole, we can accept this box as it is.

Thanks.

I thank the distinguished representative of Sweden for stating the position of her delegation and for showing flexibility. Next on my list is the distinguished representative of Pakistan. You have the floor, sir.

Thank you, Chair. Chair, we are also very satisfied with the way the box 2 currently looks. But at this moment, I've taken the floor to also share our comments on the issue of moving things from one box to another. I think the delegation of the United States has raised a very important question, something that we have heard previously as well. Chair, I would just like to recall when the GGE started in its first year, my delegation made an intervention that we think that the elements of an instrument that we are going to develop should serve three distinct purposes or could be placed in three distinct categories. And I would just like to briefly recall that. First was to reaffirm and respell existing IHL obligations as they apply to lethal autonomous weapons systems, not because these obligations are in doubt, but because we thought that their explicit articulation would carry some normative value. Second category, we thought that to further elaborate and clarify how existing IHL obligations can apply to specific characteristics of laws because of the novelty of these systems. And third, to develop additional measures wherever they are required that go beyond the existing framework. So, Chair, we have approached during the work of the GDE with this conceptual framing. And on the question of boxes, we recognize that the box structure was introduced first simply as a practical tool for organizing our discussions, and we have really appreciated its utility in this regard. However, my delegation has made it clear previously as well, and we would just like to highlight that the ultimate location of any given element within a future instrument will depend on the final nature of that instrument, of course, once we start negotiations, and not on the organizational structure, which we have adopted for the purposes of our GGE. Therefore, we have not invested significant energy in debating which concept belongs in which box at this stage. and we would just like to reserve our position on those structural questions. What matters to us at this point is the substance, ensuring that each element that we are debating is legally coherent, could be operationalized, and reflective of the mandate which you have also just read out in your intervention. Chair, on the issue of context-appropriate human judgment and control, we would like to weigh in when we go to box 3 during the course of our discussions. Thank you.

I thank the distinguished representative of Pakistan for indeed repeating what I think it was you said at our first session in 2024, when we discussed how to move forward with our work. Next on my list is the United Kingdom, to be followed by the European Union and the Russian Federation, but first the UK. Sir, you have the floor.

Thank you, Mr. Chair.

I apologise.

Thank you. To come back on a couple of the comments that we have heard in this session so far, we appreciate the clarification in relation to the vision of the document. It is really helpful for us to keep that in mind at the moment as we start to discuss what goes in the particular sections, albeit with the idea that the boxes will be removed. We did want to just note the point that the US made and agree with the point that we do see specific provisions of IHL being reflected in box 3. So in terms of the flow of the document, it would seem that we have sort of general guiding principles in box 2 as they currently sit, and then specific principles that then lead to something further in box 3. Speaking to paragraph 3 of box to where we have the IHL imposing obligations on states. We endorse that provision being in box 2. We think that it is very sensible to have that point, and I think that we made the point in the last session that it is useful to ground the concept of who the legally accountable actors are in IHL. But in that regard, and I don't want to misquote my colleague from the US, but we also heard what he was saying about how that might operate through a system of command and control. Because of course, to bridge the gap between what we have between the states being the, those that bear the obligations, the obligation to ensure respect is affected away, and that is affected through command and control. That bridges that gap into the human. We think that that might be a useful addition, and we look forward to having a look at the US's suggestions in that regard, because there could possibly be a little bit more here. Thank you, Mr. Chair.

Thank you. And indeed, the Chair is also looking forward to see the proposals from the US in writing so we can carefully look at them and see how that would, as you in a way, careful not rephrasing your words, but in a way that the flow indeed is assured. Next on my list is the European Union and then the Russian Federation, but first the EU. So you have the floor.

Thank you, Mr. Chair. Since it is the first time I am taking the floor today, I would like to thank you, Ambassador, for your tremendous efforts and excellent leadership as Chair of this Group. We appreciate your positive leadership to build the necessary momentum, also through the intersessional consultations, which has resulted in a substantial, interactive and constructive debate throughout this GGE. The European Union acknowledges the substantial contribution the GGE on LAWS has made to our common understanding of this complex topic and to finding a common ground. The EU is fully committed to deliver and fulfill the mandate of the group as soon as possible. We expect that a set of elements of an instrument without prejudging its nature and other possible measures to address lethal autonomous weapons systems will be agreed in order to move forward. We stand ready to contribute to an ambitious and substantive outcome which reflects the urgency of the matter. We believe the group is not far from reaching agreement on the elements contained in the rolling text and that they will be presented to the review conference as part of the final report. We are glad to see the elements that the European Union had proposed in the previous GGE sessions were incorporated in the rolling text. And regarding the box 2, we support reference to the need to ensure that the development and use of emerging technologies in the area of laws must be in compliance with the international law, in particular international humanitarian law. I will comment on other boxes throughout the week. Thank you very much.

Let me thank the distinguished representative of the European Union for giving her reactions to box number two. Next on my list is the Russian Federation, to be followed by Austria. But first, I pass the floor to the distinguished representative of the Russian Federation.

Thank you, Chair. First of all, We would like to say that all the five paragraphs in box 2, as they are presented by you, are acceptable for our delegation. We would be ready to agree to those paragraphs in the drafting in which they have been presented to us by you, Chair. In light of the comments that were made by the delegation of the United States of America, we would like to make a number of comments. First of all, about the boxes and moving paragraphs between the boxes. We consider these boxes to be a purely assistive tool, which was included in your document and which will not appear in the draft report of our group. As our goal is to formulate specific understandings regarding these technologies. without categorizing them into specific sections or boxes within the document itself. If this poses a problem, distinguished Chair, then we would recommend to you to not divide the text into boxes within the document. so as to remove this issue, which in our view is completely artificial. With regard to the comments on the substance now, our delegation is increasingly concerned by the situation in which certain delegations at every session bring with them new bits of text which were not discussed previously within this group and which largely make our work more difficult and make it harder to reach consensus, especially at this advanced stage of the group's work. If it is to proceed like this as we go ahead, we would like to reserve the right for our delegation to also put forward new text and new points for each of the boxes of this document, as we also have things that we would like to add. Much of what we would like to see in this document is not there, but we are prepared to show restraint in order to reach compromises and consensus on a substantive result. As we recall, most of the proposals made by the United States referred to specific guiding principles and the formulation of these principles, which were approved within the CCW. In our view, this problem could be resolved in a fairly straightforward and elegant fashion by including a separate point, which would reaffirm the understandings that are both in the guiding principles approved by States Parties to the CCW and the concepts agreed to by States Parties in the report of this group of governmental experts. In our view, this would be a fully justified approach. The results that we achieved here would be included, and these results can or could fully be approved in the context of us delivering on this mandate, the current mandate of the group of experts. As regards the new text or new text, While we do not envy your situation, Chair, if you decide to include fresh text, that will require time. Our delegation will need time in order to work through those proposals and to hash out our national position on them. By way of conclusion, distinguished Chair, As regards the proposal to include a point regarding human judgment and control, in our view, this discussion is something that awaits us when we discuss box 3, in which this matter is or can be discussed in great detail, and our delegation has a number of thoughts about this point, but we will save them for later when we actually move on to the relevant paragraph of box 3. Thank you.

I thank the distinguished representative of the Russian Federation for commenting not only on the text, but also, well, not on the text, because you that it was agreeable to you, as, of course, nothing is agreeable until everything is agreeable, but also on our way forward and how to deal more in general. I have stated at the beginning that I would appreciate if we could really focus on trying to bridge still existing divergences. And there are many ways, of course, of doing that, but indeed, we have to be careful not to overload ourselves. with, um, with proposals, uh, because as stated before, we have left 10 working days, and I would say by now, uh, we have only left nine working days, so that is clear that we have a challenge. Uh, on the other hand, uh, the Chair is fully open to constructive and creative proposals, uh, if they indeed, uh, command, um, uh, agreement among the room, uh, and not open new discussions. Let me now pass the floor to the distinguished representative of Austria. So you have the floor.

Thank you very much. Coming to box 2, we originally did not intend to take the floor on this box since we thought that the current proposal works. In this context also, I wanted to point to our intervention during the informal consultations where we said that the last para in box 2 was a very elegant and very good bridge between box 2 and box 3 by introducing the concept of context appropriate human judgment and control. Since then, we have looked into your proposal and the way how you fix this by taking this para and putting it in box 3. works for us as well. It takes out this elegant path, but it makes it more stringent and more direct and also solves some other issues. So we could have worked with that, even though we would have lost something that we very much liked in box 2, insofar as it has linked the general principles with everything that follows. And that was our understanding of the text. When it comes now to the purposes of the boxes of box 2 and 3, We also believe that in the end this will be something that is introduced into one text. We don't express ourselves at this point on preamble or paras and how it will link up. This is probably for a later stage to be discussed and also to format it. The right place to discuss the concept of context-appropriate human judgment and control would definitely be box 3. Insofar as we can have this discussion, we also appreciate the proposals by the US by having a discussion and looking into possible solutions. But mostly, the current formulation that has been accepted by many delegations here in box 2 and now in box 3, as far as we can see, is the best solution at this point in time. The other proposals about adding other elements in box 2, many of them are quite new. It has also been very hard for us to grasp them at this point in time. I mean, we would have to look at them in writing. But at the same time, I have to point out that we had long discussions on box two and also on the parts that have now moved into box two. It has been the most advanced and most agreed boxes of all of them for a very long time, now shuffling new elements into and around, and some of them which are probably also better in placed in other boxes like the accountability issues. We also don't completely understand why we would take the para 3, which is of the 2019 report and mix it with guiding principle D. It's very random, I have to say, and does not really give this linkage that is supposedly it's supposed to bring to the text. We don't really understand that at this point in time, so we would rather asked to reframe from adding other elements here than the one that you have proposed and mixing up the paras with new text as well. We all know the guiding principles, we all know the 2019 report, we have discussed them in different parts of the text in the past two years, it's not necessary to shuffle them around a new and a new and a new in different places and different boxes. That's what we have done in the previous two years and now we are fixing what is in front of us for a final outcome of this group. Thank you.

Let me thank the distinguished representative of Austria for also reacting to proposals made earlier this afternoon. Next on my list is Spain, to be followed by Peru. But let me first pass the floor to the distinguished representative of Spain. You have the floor.

Muchas gracias.

Thank you very much, Chair. Allow me to make the following additional comments on top of what the EU said. First of all, we won't go into the proposals for moving around the paragraphs between the different boxes, as we understand that the box structure was a tool for facilitating our work and our discussion theme by theme, and we're very grateful for that. At this stage of the process, with only two more sessions to go and on day 2 of this current session, we do not believe it is appropriate to include new language or new paragraphs, because that would move us away from delivering on our mandate. Secondly, for my delegation, it is important to encompass references to international humanitarian law and the obligations it entails and the framework for application. Therefore, we believe that it is important to include this box and to include a specific scope. We also welcome the new paragraph 3 on the IHL imposing obligations on states, not on machines. That might be stating the obvious, but that doesn't mean it's redundant here. We also believe that it is useful to keep the reference to the Martens clause. And fourthly, with a view to limiting human judgment and control to being context appropriate, we felt it was appropriate to include it in this box. That was our view in the past, but we would not have any objection to moving it, as you did, to box 3. Finally, allow me to convey to you my delegation's support. We fully support box 2 as it is, as you have presented it to us. Thank you.

Let me thank the distinguished representative of Spain for stating the position of her delegation and sharing views with regard to the structure of the document. Next on my list is Peru and to be followed by Lesotho. Yes, but let me first pass the floor to distinguished representative of Peru. You have the floor.

Thank you, Chair. As this is the first time that my delegation has taken the floor, I would like to begin by commending you on your work as the chair of this group of governmental experts. As has been said by previous speakers over the course of today, we would also like to say that although the rolling text remains a work in progress, it can be considered as a sufficient basis for us to accomplish the mandate of this group on. With regard to box 2, I'd like to make three comments. Firstly, I'd like to say that we welcome the effort you have made to maintain this text, which reflects an important reaffirmation of international humanitarian law and which, as you yourself said, received broad support during the last session of 2025. We have also taken note of the relocation of paragraph 3, which addressed human control. And the new paragraph 4, which includes the references to development and use, in keeping with Article 36 of Additional Protocol I. And finally, I would also like to add my delegation's voice to what has been said about the structure and the placement of the various elements, the various set of elements that we're discussing in the different boxes. And although I think it is useful to start to have this conversation, for now, the system that you have been using has shown to be very useful for structuring our work and for giving a clear outline to our discussions. Thank you very much.

I thank the distinguished representative of Peru for sharing the views of his delegation, also with regard to the structure of the document. Next on my list is the distinguished representative of Lesotho. You have the floor.

As this is our first time taking the floor, allow us to appreciate your continued leadership and commend the inclusive approach taken in developing the rolling text. We appreciate that this text reflects constructive contributions from many delegations and represents important progress in our collective work. We remain ready and committed to completing this mandate's work and support moving towards negotiating on the basis of the rolling text. Now moving to box 2 and 3. We find it commendable that both box 2 and box 3 try to ensure the applicability of the principles of IHL to laws. We are glad that box 2, paragraph 3, places accountability on states and individuals and not on machines. Box 1, paragraph 1 of box 3, keeps with the spirit of our umbrella convention, CCW, which restrict or prohibit the use of weapons that are deemed to be extremely injurious or non-discriminatory in effects. And box-- paragraph 2-- box 3 talks about the principles of anticipation. Paragraph 3 talks about discrimination between military objectives and civilian objectives in targeting. So we find this to be very commendable, that the principles of IHL are now covered are now being applied to laws. We strongly support the principles in box 3, paragraph 6, that context, appropriate human judgment and control is essential for compliance with international law. This represents the cornerstone of the ruling text and must be preserved. The measures outlined in paragraph 7 operationalize these principles and reflect the key components of meaningful human control. Human operators must remain accountable for the operation and effect of weapon systems, which require users to understand the system and apply context-specific decision-making and.

Sorry, your microphone dropped. I think you have to once more say the last few lines of your intervention.

Thank you, Chair. Thank you, Chair. My last sentences were the measures outlined in paragraph 7 operationalize the principles and reflect the key components of meaningful human control. Human operators must remain accountable for the operations and effects of the weapon systems, which require users to understand the systems and apply context-specific decision-making and judgment. Thank you, Chair.

And I thank you for also sharing already your views with regard to box number three. It would have been an excellent transition, but I think we have still a few more requests for the floor at this moment coming from Switzerland, Portugal, and Utrecht University. But let me first pass the floor to the distinguished representative of Switzerland.

Thank you, Mr. Chair. We can be brief. In light of the discussion this afternoon, like others, we think it could be wise not to present a great number of new proposals late in the process, and our impression is that this could be a principle to observe throughout the text. But here in box 2 in particular, since box 2 has generally been characterized by many as in good shape, and in the view of our delegation, this is very close to the finish line. So continue the text to support the text as it is. Like others, we see these elements as, yes, a preambular type of elements, but also we should not put it in preambular type of language. On moving things from one box to another, These can be useful, particularly we see the logic of in moving paragraph three to box two, underscoring that legal obligations rest with states and individuals, not machines. And we also support moving the paragraph on context appropriate human judgment and control to box three, paragraph six. But for us, like for other delegations, the boxes are simply for practical reasons, for structuring our discussions and not for structuring a future instrument at this stage. So we would hope that we wouldn't spend too much time on that issue. Thank you so much. That's all from us.

I thank the distinguished representative of Switzerland. Let me now pass the floor to the distinguished representative of Portugal. You have the floor, sir.

Thank you, Chair. I'll also be very brief. I'd just like to say that my delegation was ready and continues to be ready to accept the tasks that you have proposed as is. My second comment is that, of course, and it was already mentioned before, namely by Switzerland just now, that although, of course, structure and systematization is important in our exercise, trying to move paragraphs from one box to the other is of limited normative effect. We would advise against embreanding such an exercise, which will only prolong our discussions without great significative effect. Finally, we would also advise against trying to introduce new language. We heard interesting proposals. We are not denying that many of them perhaps could be useful, but at this time that would demand a new discussion, a reopening to boxes that are, in our view, already ripe and merits the consensus of the majority. or the agreement of at least of the vast majority of delegations. So and to conclude, we just like to once again support your text and just say that we would very much like to move forward with our discussion. Thank you.

I thank the distinguished representative of Portugal, and I would like to pass the floor to the distinguished representative of Utrecht University. You have the floor.

Thank you, Chair. I'm again delivering this intervention on behalf of the delegation of the Asia Pacific Institute for Law and Security and my own Utrecht University. Though we don't see the need to weigh in on the moving of the text up to box one, as other delegations have mentioned. We would like to briefly address the terminology used in box 2, paragraph 4, specifically the reference to the development and use language. Not wanting to complicate matters unnecessarily by introducing new ideas at this point, as the last two distinguished representatives from Switzerland and Portugal have mentioned, And offered in the spirit of an element to consider, we propose that in box two, paragraph four, development and use be changed to life cycle. The text would then read, IHL applies fully to the life cycle of laws. We have a similar suggestion related to text in box three, para six, but we return to that later in the week. For one, our proposed change brings cohesion across other areas of the text. The term life cycle is already used in boxes four and five. More substantively, we respectfully suggest that a reference to the life cycle would more explicitly encompass stages of planning and designing, as well as developing and using from a technical perspective, echoing a life cycle approach of the system and tracking largely across the language in Article 36 of AP1 addressed to high contracting parties in the study, development, acquisition, or development of a new weapons means or method of warfare. Clarifying box 2, paragraph 4, as well as other references throughout the text to reflect all phases of the life cycle would strengthen the text and ensure consistency with IHL. It would additionally reinforce the principles that responsibility and accountability as addressed in box 5 of the rolling text cannot be deferred to the point of battlefield deployment and above all makes the text internally consistent. I will also offer these comments to the Chair and your team after the intervention. Thank you, Chair.

Let me thank the distinguished representative of UTech University. I must say we are certainly reflecting on making the wording consistent throughout the text, so we welcome also your intervention with regard to that. I'm looking around whether there is any further appetite for interventions or that there's more appetite for going to the reception. I am afraid that as it then goes, we have still somebody who has appetite for an intervention, and of course, you are fully entitled to that. Let me pass the floor to the distinguished representative of India. You have the floor.

Thank you, Mr. Chair, for giving me the floor. And based on the discussions we had, I did want to make a few comments, but I'd be very brief and not extend discussion on box 2. First of all, we want to clarify that we can accept the box 2 as it exists. We see the value of this box in reiterating and reaffirming the core principles of IHL and our previously agreed general understandings of the group. Given the views that we have heard, I want to take opportunity to underline our views on the term control and human judgment. We agree that the terms control and human judgment are essential for IHL compliance. This has been agreed to by the group in 2023 and 2019, respectively. The group has over the years been able to attach a meaning to these terms, which have converging interpretations, if not a common one. Further, my delegation supports the view that the exercise of human judgment should take into account context-related details such as the operational context, the environment, and the characteristics or capabilities of the weapon system. An exhaustive list of these factors, however, is difficult to draw down. These factors are also dynamic, varying according to the context. We have heard different views on whether we can create a common and an objective benchmark around these terms. such that it can be presented as a specific obligation or an actionable measure. However, we believe that this term can work together as a general or a conceptual guidance. It is in this context that we believe that box 2 might serve to be the right place to enshrine the essentiality of control and human judgment. My delegation is willing to use the term context appropriate human judgment and control. However, we note that some delegations had concerns with regard to the use of this very specific term. Our view is that in order to accommodate those concerns, we should unpack the term if that helps. Our suggestion would be to use the following formulation that control and human judgment should take into account contextual aspects, which emphasizes the need on adapting human judgment with context relevant details. This term will work best when it acts as a conceptual guidance or philosophy, which is best reflected by adding to this very specific section. We would be willing to, of course, discuss this further with you and interested delegations on the exact formulation and its very placement. I thank you, Mr. Chair.

Let me thank the distinguished representative of India for sharing his comments, also with regard to the acceptability, with regard of box 2, of course, but also looking forward already to the discussion on context appropriate human judgment and control, and for coming up with a proposal for further discussion. Certainly interested, and we will see how we can have that further discussed. I'm also looking a bit at the time, um, I, I am a little bit hesitant in now opening the discussion on, on box three, uh, because that will, uh… take quite some time, I think, and then we can only listen to perhaps one or two, or not even, and that would, I think, not really facilitate the debate, because I would like to hear a flow of reactions to what we have proposed there. So, some of you have already anticipated that, and I very much appreciate that, of course. Looking back at box number two, I think it has been voiced by a number of delegations The Chair and his team, we were of the opinion that this was a box, if there's a box that is close to a preliminary agreement, consensus, that this was certainly such a box. But of course, that is always an assumption and that assumption has to be tested. Well, we tested it. I think we have not succeeded in passing On the other hand, I think there's also a strong feeling among the room that we should not change too much to box 2, also taking into account that there's still a lot of work to be done on especially box number 3, and we have also dedicated time for that both tomorrow and later in the week. Now, looking at my list, I have one further more request for a right of reply. Let me underline again what I said at the beginning of our session today, that I do hope that we can focus this week on what we have been tasked to do. I understand this delegation wants to give a very brief reply to remarks made earlier today, so I will now give the floor to the distinguished representative of Israel. You have the floor, sir.

Dear Chair. It was not our intention to engage in a politically motivated discussion, and it is not our intention to continue doing so. But some grave accusations heard in this room cannot remain unanswered. We categorically reject these accusations. Israel is engaged in an ongoing armed conflict against the Iranian regime, which has repeatedly attacked and continues to act in aggression against Israel, both directly and through its proxies, while developing a military nuclear program and an arsenal of thousands of ballistic missiles. The current operation was initiated to remove existential threats posed by the Iranian regime. Israel was compelled to act before it would be too late to effectively counter these threats. Israel will continue to defend its citizens while acting in accordance with international law. Mr. Chair, my delegation is ready to engage actively and constructively during this week in order to reach a consensus on the important issues this forum is meant to discuss. I thank you.

I thank the distinguished representative of Israel. I now have, before adjourning the meeting, I have one Yes. I would like to kindly inform you of the side event organized by the Stockholm International Peace Research Institute, SIPRI, which will be held tomorrow morning from 9:00 until 9:50 in room 12 of building A of the Palais des Nations. This event will introduce a new report from SIPRI, which examines the intersection between military procurement and responsible AI. Let me repeat, this side event organized by SIPRI will be held tomorrow morning from 9:00 to 9:50 in room 12. And let me also add a brief logistical point. It's possible that some of you are using a non-permanent batch, and then you may face difficulties for entering building A. In such a case, please kindly use entrance A from Alpha 17. I repeat, Alpha 17, and use the interphone available to inform security that you are willing and looking forward to participate in the side event organized by SIPRI. They should then let you in. They must let you in, I would say. And you will also find more detailed information on the webpage of the 2026 session of the GGE on laws. And that concludes then our discussions for today. And of course, we have one more job to perform, and that is the reception that will take place outside. I look forward to seeing you there and to continue our informal discussion then, or discussion in an informal mode. Have a great evening, and see you back tomorrow morning at 10:00 in this room. The meeting stands adjourned.