From technology to law: The case of sea drones
In the current era, technology plays a significant role in armed conflicts.Footnote 1 In 2022, the North Atlantic Treaty Organization (NATO) stated that technological primacy had a direct impact on battlefield success,Footnote 2 and highlighted both the positive and negative aspects of emerging and disruptive technologies (EDTs).Footnote 3 The Organization also underlined that EDTs not only alter the character of conflict but have also acquired greater strategic significance, thereby becoming the “key arenas of global competition”.Footnote 4
The ongoing developments in the naval warfare between the Russian Federation and Ukraine in the Black Sea since 2022 represent a primary example of how EDTs impact the dynamics of a conflict, creating naval asymmetry and raising ethical questions.Footnote 5 There are, undoubtedly, multiple examples of new technological solutions employed by these two States, but it is the Ukrainian sea drone attacks against Russian armed forces over the last few years that have garnered worldwide interest.
Based on incidents reported in publicly accessible sources, several significant engagements involving sea drones can be identified. Among these are the October 2022 Sevastopol attack, where Ukrainian sea drones are reported to have damaged the Russian-flagged Ivan Golubets (Natya-class minesweeper) and the Admiral Makarov (frigate).Footnote 6 In 2024, Ukraine is reported to have launched a series of attacks deploying the MAGURA V5 (Maritime Autonomous Guard Unmanned Robotic Apparatus V-type) sea drone, resulting in the sinking of vessels such as the Ivanovets (Tarantul-III corvette), Tsezar Kunikov (landing ship) and Sergey Kotov (patrol ship).Footnote 7 In 2025, media reported the first-of-their-kind aerial shoot-downs of a Russian Mi-8 helicopter and a Su-30SM fighter jet by MAGURA types V5 and V7 respectively.Footnote 8
A close examination of these developments reveals that sea drones not only shape the military landscape but also challenge the legal framework governing the conduct of hostilities during an international armed conflict. The deployment of sea drones raises multiple legal questions, but the most fundamental one relates to their legal status and categorization as either a warship or a means of warfare under the law of the sea and international humanitarian law (IHL), notably the law of naval warfare, and the different legal consequences of the two categorizations.
To explore this issue, the remainder of this paper proceeds as follows. The first section classifies sea drones under the “warship test” – i.e., it assesses their characteristics against the requirements derived from the definition of warship codified in Article 29 of the 1982 United Nations Convention on the Law of the Sea (UNCLOS).Footnote 9 The second section explains the alternative legal classification of sea drones as a means of warfare under Articles 35–36 of Additional Protocol I to the Geneva Conventions (AP I).Footnote 10 The third section explains the legal consequences associated with classifying sea drones under one or the other category. In conclusion, it is argued that the warship crewing requirement, as outlined in UNCLOS Article 29, is not merely a technicality; it is a decisive factor that distinguishes a warship from a weapons system, and also carries significant implications for the humanitarian aspects of conflicts.
What is a sea drone? A brief look at terminology
Given the novel nature of sea drones and their emerging status, there is no definition of a sea drone in either legal or technological terms. Before turning to the terminology used in the legal scholarship, it is therefore necessary to outline the basic technological categories within which sea drones are situated. An “unmanned system” is the broadest category and refers solely to the absence of humans on board. On the other hand, an “autonomous system” answers the question of who – i.e., a human or an artificial intelligence (AI) – undertakes decisions; thus, it describes the operational decision-making process.Footnote 11 A fully unmanned system may therefore be remotely operated and not autonomous,Footnote 12 and conversely, a fully autonomous one may still have humans on board, such as weapon operators or other personnel whose tasks are unrelated to operational decision-making. Naval systems are simply those used for military purposes at sea.Footnote 13
The limited legal scholarship employs various terms such as “unmanned maritime systems”,Footnote 14 “unmanned maritime vehicles”,Footnote 15 “unmanned naval systems”Footnote 16 and “maritime autonomous vehicles”.Footnote 17 Even though these terms are not fully synonymous, they describe overlapping categories that are (inconsistently) used in the literature to capture similar technologies. Most legal scholars address the broader phenomenon of drones in the context of land and aerial warfare, referring to them also as “autonomous weapons systems”Footnote 18 and “lethal autonomous weapons systems”.Footnote 19 In a naval context, the concept of “lethal autonomous maritime systems” has also appeared.Footnote 20 Terms such as “sea drones” and “maritime drones” are often used interchangeably to describe a comparable kind of machine, usually referring to military technology.Footnote 21 This article focuses on the deployment of sea drones in warfare, leaving aside their use in civil and commercial contexts.Footnote 22
For these reasons, given the lack of a widely accepted definition in a military context, a sea drone will be understood here as a machine operating in a maritime environment, equipped with a system that allows it to operate with limited or no human involvement in the decision-making process (human-in/-on/-off-the-loop),Footnote 23 and without human presence on board (unmanned/uncrewed).
This definition encompasses systems with varying degrees of autonomy. A sea drone may be fully autonomous in the decision-making process (human-off-the-loop),Footnote 24 may operate with human oversight over autonomous decisions, where a human enjoys veto power (human-on-the-loop), or may be fully dependent on the operator’s decisions (human-in-the-loop). The manning/crewing levels refer only to the physical presence of humans on board. Although sea drones are generally presumed to be unmanned by definition, the meaning of “manned” craft under the current law is disputed, as will be discussed further below. Conceptually, a sea drone may have no human on board at all (fully unmanned), a limited on-board crew, e.g. to operate weapons (partially manned), or a regular crew on board (fully manned). Consistent with the terminology used in the existing scholarship on “unmanned” crafts, the present analysis focuses on unmanned variants of sea drones.
While this definition focuses on characteristics relevant to a sea drone’s legal status rather than its specific operational role, the present analysis centres on those sea drones that are capable of delivering force – i.e., those that are designed or equipped to cause injury, damage or death. The broad scope of the proposed definition is intentional, aiming to encompass a wide range of machine types. This approach is particularly significant given the heterogeneous and evolving nature of technology, and reflects an intent to avoid excluding any relevant systems whose absence could undermine the legal analysis. The primary objective of this definition is to identify the functional characteristics that matter for legal classification, not to provide a technical taxonomy. Additionally, in the absence of universally accepted definitions, it underscores that the two characteristics discussed above – namely, autonomy and manning – are the sole stable conceptual anchors across all variants.
These considerations are not purely technical or theoretical; they have practical implications for the legal analysis of a sea drone’s status. Only a detailed understanding of the analyzed craft can lead to a comprehensive legal classification and the assignment of the corresponding consequences. The following sections provide a detailed analysis of the status of a sea drone under international law.
The sea drone as a warship
Warships represent one of the most enduring categories of seagoing vessels, used throughout the centuries to secure maritime dominance. The contemporary warship definition is codified in UNCLOS, Article 29 of which states thatFootnote 25
[a] “warship” means a ship belonging to the armed forces of a State bearing the external marks distinguishing such ships of its nationality, under the command of an officer duly commissioned by the government of the State and whose name appears in the appropriate service list or its equivalent, and manned by a crew which is under regular armed forces discipline.
The following historical overview explains how this definition emerged and why its elements matter for assessing whether sea drones could fall within this category. Warships, understood broadly as State vessels through the lens of a functional link, were, for the first time, legally distinguished from those operating as “licensed plunderers”Footnote 26 – privateers – in the 1856 Paris Declaration Respecting Maritime Law.Footnote 27 Building on this fundamental separation, the 1907 Hague Convention (VII) relating to the Conversion of Merchant Ships into War-Ships provided the key requirements for classification as a warship, including that it is under State authority, bears external identification, follows a command structure and is manned by a disciplined crew.Footnote 28 However, it was not until the 1948 Geneva Convention on the High Seas that a single-line warship definition was drafted.Footnote 29
This definition established the basis for the current formulation in UNCLOS Article 29, as presented above. Based on this contemporary definition, a craft must have four basic characteristics to be classified as a warship. First, it must be a ship; second, it must form a part of the State armed forces, bearing marks of its nationality; third, a duly commissioned officer must be in command; and fourth, it must be manned by a disciplined crew. The following sections test sea drone characteristics against those established by UNCLOS Article 29.
State practice and military manuals
An analysis of State practice and military manuals sheds light on how States perceive this emerging category of crafts; nonetheless, evaluating State practice remains challenging due to limited public access to naval registers and the lack of explicit classifications for such crafts. A few available examples demonstrate distinct approaches to the latter issue.Footnote 30 The US Commander’s Handbook on the Law of Naval Operations (US Commander’s Handbook), while analyzing the warship status, explicitly states that “[t]here is no requirement [that] the commanding officer or crew be physically on board the warship. Warships may be remotely commanded, crewed, and operated.”Footnote 31 The 2025 Newport Manual on the Law of Naval Warfare (Newport Manual) reaffirms this view by indicating that “[w]arships may … be autonomous or remotely commanded, crewed, and operated by personnel ashore or on board a manned platform”.Footnote 32
On the other hand, the Australian Naval Classification Manual classifies remote and autonomous systems as “naval vessels”Footnote 33 – a broader category than “warship”, which indicates that a vessel is a military-operated ship, regardless of whether it is designed for combat. In a similar tone, the US Commander’s Handbook explains the “unmanned system status” by stating that
[i]n all cases, U.S. Navy [unmanned systems] … [w]hen flagged as a ship … may exercise the navigational rights and freedoms and other internationally lawful uses of the seas related to those freedoms. Unmanned systems may be designated as USS [a Navy ship] if they are under the command of a commissioned officer and manned by a crew under regular armed forces discipline, by remote or other means.Footnote 34
Conversely, Ukraine, when referring to the MAGURA sea drones discussed earlier in this article, typically designates them as “naval drones”, “maritime strike drones” or “weapons”. There does not seem to be any publicly available evidence indicating that Ukraine has officially classified them as “warships”, or that it has even registered them as ships at any point at all.Footnote 35
The Newport Manual expresses an interesting view in this regard:
Unmanned Maritime Systems (UMS) are prevalent in armed conflict at sea. For example, since the escalation of the Russia-Ukraine War in 2022, both States have employed remotely controlled, explosive-laden, expendable UMS to conduct attacks. These UMSs have been used successfully to conduct attacks on enemy ships at sea and in port, as well as on lines of communication (e.g., bridges). Neither side has indicated whether they have designated their UMSs as vessels or consider them to be a weapon system or vessel or both. According to the present Manual, such systems qualify as weapons rather than as UMS that could be classified as warships.Footnote 36
This approach appears to introduce operational characteristics such as the “remotely controlled, explosive-laden, expendable” nature of certain systems as determinative factors in their legal classification. Yet these considerations do not stem from UNCLOS Article 29, or from established doctrine on warship status, and the Newport Manual does not clarify whether such operational features form the basis of its own classification.
This demonstrates a broader lack of coherence in current State and doctrinal approaches to unmanned maritime systems. Thus, while US practice suggests that unmanned systems may qualify as warships, the other provided examples – such as those of Australia and Ukraine – do not yet allow conclusions about broader State practice, necessitating a doctrinal analysis of UNCLOS Article 29.
Ship requirement
Public international law does not offer a legal definition of a “vessel”. Terms such as “vessel”, “ship” or “boat” are also often used interchangeably to describe the same concept. Linguistically, a vessel is defined as “a ship or large boat”, while a ship is “a large boat used for traveling long distances over the sea”.Footnote 37 A boat, on the other hand, is simply “a vehicle of any size used for traveling on water”.Footnote 38 Furthermore, the law of the sea and international maritime law conventions define these terms differently.Footnote 39 However, none of these definitions refer to the operational mode – i.e., who is responsible for the decision-making process – or require human presence on board to classify a craft as a vessel. The scholarship also generally agrees that autonomy and manning levels do not affect the ship’s classification.Footnote 40 For these reasons, a sea drone, independently of the human-in/-on/-off-the-loop configuration, and with or without human presence on board, can be classified as a vessel: it is a waterborne craft designed for navigation and operation at sea, and thus, it complies with the first UNCLOS Article 29 warship requirement. Limited available State practice also supports this interpretation, as several States classify unmanned surface craft as vessels for regulatory purposes.Footnote 41
External marks and attribution requirements
The integration of sea drones into States’ armed forces is becoming more common. Examples from States such as Ukraine, Iran, the United States, the United Kingdom and China demonstrate that maritime unmanned and autonomous systems are now a part of not only navies but also broader military forces.Footnote 42 Thus, a sea drone may belong to the State’s armed forces in the understanding of UNCLOS Article 29, even though it is not exactly clear whether, for example, the State’s ownership, control or responsibility is required by this provision.Footnote 43 Furthermore, if a State considers its sea drone to be a vessel, it should flag the drone under UNCLOS Article 91.Footnote 44 Thus, there are no technical or legal obstacles preventing sea drones from bearing the external marks distinguishing them by their nationality; they may, similar to conventional warships, carry a visible naval ensign.Footnote 45
Command and official listing requirements
As stated by UNCLOS Article 29, a warship must be “under the command of an officer duly commissioned by the government of the State and whose name appears in the appropriate service list or its equivalent”. Two important questions emerge in this context regarding sea drone compliance. One difficulty lies in the fact that neither UNCLOS nor any other international regulation explains what the term “under the command” encompasses; another concern relates to whether AI could qualify as a commanding officer, in the case of sea drones with human-off-the-loop.
To begin with, as noted above, international law does not define the term “under the command” – or even the term “command”. There are also various definitions of this concept present in the military scholarship.Footnote 46 The most basic one describes command as “act[ing] on a will”,Footnote 47 while others add that such a will must be necessary to “accomplish the mission”.Footnote 48 On the other hand, an ordinary meaning of the term “command” encompasses ordering someone in a “forceful” and “official way” to undertake conduct.Footnote 49 None of the provided definitions invoke the necessity of the shared situs – i.e., that a person issuing the command needs to be physically located in the same place as the subject of the command. In light of the foregoing, an unmanned sea drone with human-in and -on-the-loop can be “under the command” of a remote operator in accordance with the analyzed provision.
The second question of whether AI can be regarded as an officer duly commissioned by the State presents a particularly challenging issue. Providing an answer to this question remains especially important within the context of fully autonomous and unmanned sea drones (operating with human-off-the-loop). Although such a scenario is currently unlikely, given that publicly known technological development primarily emphasizes fully unmanned systems with varying levels of operational autonomy but with human-in or on-the-loop, it cannot be excluded that classified programmes may follow different trajectories. Hence, the potential future significance of this question remains noteworthy. In such a case, an AI would need to be granted a formal military State commission, registered in the official records, to be regarded as an officer. Yet, this would not be possible without the AI achieving a human-like status first (i.e., personhoodFootnote 50), given that – as of now – only persons can be members of the armed forces. Thus, since a philosophical discussion on the nature of robots and AI exceeds the scope of this article, it must be concluded that a fully autonomous sea drone with human-off-the-loop cannot, at this point of unresolved scholarly debate on AI personhood,Footnote 51 comply with the mentioned requisites derived from UNCLOS Article 29 on the notion of command.
Manning and discipline requirements
The last requirement – being manned by a crew under regular armed forces discipline – is the most challenging part of the warship test under UNCLOS Article 29. It requires addressing two questions: whether a sea drone can be “remotely manned”Footnote 52 and therefore classified as a warship, and the scope of the term “crew under regular armed forces discipline”.
Firstly, it is worth noting that manning is one of the key factors that enable a vessel to achieve warship status. In other words, while the craft’s status as a vessel is not affected by the presence or absence of a crew on board, UNCLOS Article 29 explicitly requires the physical presence of a crew to categorize a vessel as a warship. The construction of this provision is rooted in the prohibition of privateering, mentioned previously, and the need to distinguish State vessels from those operated by private individuals during naval hostilities.Footnote 53 It derives from the presupposition that a warship is “an expression of the sovereignty of the State whose flag it flies”.Footnote 54 Consequently, a warship enjoys certain rights and obligations, absent in the context of non-State vessels,Footnote 55 which are further discussed in the next section.
The “regular armed forces discipline” requirement is a concern similar to the case of the commanding officer, whose name should appear on the appropriate service list or its equivalent, as addressed above. For the same reasons, persons involved in a sea drone’s operations, to be considered crew members under UNCLOS Article 29, must be members of the flag State’s armed forces. The engagement of only private contractors or civilians in the operation of sea drones, without any disciplined crew, does not comply with the established provisions. The presence of a disciplined crew on board presumes service within a clear hierarchical order and chain of command.
Consequently, under the current regulations, the physical presence of crew on board is a decisive factor distinguishing a warship from a civilian vessel or undefined floating craft.Footnote 56 Thus, fully unmanned sea drones, even with a certain level of remote human involvement at the operational level (e.g., with a shore-based human-in and -on-the-loop remote operator), cannot comply with the requirements of UNCLOS Article 29 on manning.Footnote 57
To summarize, fully unmanned sea drones (which are the focus of the present analysis), independent from human involvement at the operational level, fail to pass the UNCLOS Article 29 warship test. Even if they fulfil certain criteria, such as being flagged vessels, and some interpretative flexibility allows for accommodating a shore-based remote operator as the commanding officer, the manning requirement is not met. For these reasons, an alternative classification of sea drones is necessary and can be found within the international humanitarian law regulatory framework. The following section analyzes this alternative.
If not a warship, then what? Sea drones as weapons
As established above, manning by a crew under regular armed forces discipline is what could make a sea drone a warship. In the absence of manning, sea drones capable of projecting force are closer in nature to means of (naval) warfare,Footnote 58 which are understood as weapons and weapons systems deployed during an armed conflict.Footnote 59 Military doctrine and scholarship consider means of warfare in a broad sense as devices, substances, systems, munitions and other types of equipment designed to cause injury or damage to an opposing force during an armed conflict.Footnote 60 Naval mines, torpedoes, missiles and other projectiles are the most common types of naval means of warfare.Footnote 61
This article argues that certain sea drones share certain functional and operational characteristics with naval means of warfare, especially naval mines, as explained further in this section. A brief comparison with torpedoes is also presented; however, since the legal distinction between torpedoes and mines is rather blurry,Footnote 62 this comparison provides limited analytical input. Sea drones are not, however, comparable to missiles and other projectiles, given that the latter are single-use munitions designed to be launched and to travel on a fixed trajectory toward a target,Footnote 63 which is not the case with sea drones. Sea drones also intersect conceptually with the broader notion of autonomous weapons systems, a topic that is increasingly discussed in scholarship. However, as the scope of this section is limited to naval means of warfare, a broader debate on autonomous weapons systems is left for future work.Footnote 64
The features of naval mines, as presented below, correspond to the emerging category of sea drones, whose operational profile closely aligns with many functions traditionally associated with naval mines. Multiple scenarios might be developed to demonstrate the operational similarities between the two. For example, an armed and unmanned sea drone may be left to drift with water currents, triggering attack or detonation based on the impulses received from its in-built sensors; sea drones may also remain free-floating in an uncontrolled manner as post-hostilities threats, similar to naval mines.
On the other hand, the functional characteristics of sea drones, such as mobility, speed, strike capacity and precision, support the analogy with torpedoes, which are self-propelled underwater munitions with an explosive warhead designed to target other vessels.
Therefore, this section establishes an analogy between sea drones and these two types of naval means of warfare (i.e., naval mines and torpedoes), considering that naval mines and torpedoes serve as the most relevant historical and legal precedents for maritime autonomous and unmanned systems, despite not having identical characteristics with sea drones.Footnote 65 Accordingly, the international framework applicable to naval mines and torpedoes is analyzed below to reach pertinent conclusions.
A wide range of naval mines exists in the current stage of technological development, from traditional contact mines to smart and missile-age mines. Modern smart mines, such as the Turkish Malaman, are reported to be able to identify targets, categorize them, and trigger detonation at the strategically optimal moment.Footnote 66 All of this is possible, as reported, due to the complex algorithms that these mines are programmed with, which allow them to analyze data in real time and make autonomous decisions.Footnote 67 Remote operators may activate and deactivate them from a distance, and they are also equipped with a self-neutralization mechanism. From a functional standpoint, naval mines are unmanned devices deployed in an aquatic environment. Their tasks are to deter, damage or destroy other vessels, and they can also establish zones of denial or exclusion at sea (e.g., naval blockadesFootnote 68). Sea drones may be deployed for similar purposes.
To date, however, there is no legal definition of naval mines or torpedoes at the international level,Footnote 69 and the sole international convention addressing mine and torpedo warfare – the 1907 Hague Convention (VIII) relative to the Laying of Automatic Submarine Contact Mines (Hague Convention VIII)Footnote 70 – does not provide any definition for them.Footnote 71 Adopted in the aftermath of the Russo-Japanese War of 1904–05,Footnote 72 Hague Convention VIII aimed to diminish the negative impact of war while enhancing the safety and security of navigation despite it.Footnote 73 It provides a catalogue of restrictions and prohibitions for automatic contact mines (those that explode upon contact with a vessel), the primary type of mines at the time of its adoption. Firstly, Hague Convention VIII prohibits laying unanchored, free-floating contact mines, unless their construction allows them to become harmless within a maximum of one hour after losing control over them.Footnote 74 Secondly, it bans the usage of anchored automatic contact mines which do not become harmless immediately after breaking loose from their tethering.Footnote 75 Regarding torpedoes, Hague Convention VIII mentions them only once, stating that they must become harmless after missing their target.Footnote 76 Finally, it obligates Neutral Powers, belligerent States and Contracting Parties to provide navigational safety through various means to the fullest extent possible in the zones where mines are laid.Footnote 77
Over the years, Hague Convention VIII has diminished in legal significance, mainly due to technological developments and the deployment of modern types of naval mines. Nevertheless, the core principles it created have become customary international law,Footnote 78 providing guiding principles for the review of naval mines.Footnote 79
The analogy between sea drones and naval mines/torpedoes bears concrete regulatory lessons for the former.Footnote 80 In broad terms, sea drones considered as a new means of warfare would fall within the scope of AP I Article 36 on weapons review obligations, requiring States to assess their compliance with international law prior to deployment. Also, belligerent States should, as a matter of lege ferenda, ensure navigational safety,Footnote 81 mainly of civilian and neutral vessels, in sea drones’ operational regions.
This latter task can be accomplished to varying degrees depending on the sea drone’s configuration. In the case of fully unmanned sea drones with human-in-the-loop, a remote operator should be able to steer the drone away from neutral and civilian vessels, diminishing the risk of indiscriminate attacks. In a scenario involving a sea drone with human-on-the-loop, the danger that a sea drone will drift into neutral waters or engage with neutral or civilian vessels is greater due to possible delays in the remote operator’s intervention. Finally, fully autonomous sea drones, which operate with human-off-the-loop, may cause the most severe difficulties and risks. The absence of real-time human oversight and the unpredictability of AI-operated sea drones (e.g., the black box problemFootnote 82) may result in violations of both treaty and customary principles governing naval warfare (e.g., indiscriminate attacks, lack of precaution, or attacks violating the principle of proportionality, to name just a fewFootnote 83).
Also, similar to naval mines, as a matter of lege ferenda, fully unmanned sea drones should self-neutralize within one hour after the remote operator loses control (human-in-the-loop) or oversight (human-on-the-loop) over them. In the case of fully autonomous and unmanned sea drones, a similar requirement should be enforced after they lose contact with their operational headquarters. In the event that sea drones are deployed for one-way attacksFootnote 84 (i.e., missions in which they are intended to be expended upon impact, similar to torpedoes), they should self-neutralize after missing their target.
To summarize, a complex characteristic of sea drones, namely the combination of being unmanned and having varying degrees of human involvement at the operational level, challenges the existing legal framework on means of naval warfare. Nevertheless, as demonstrated in this section, multiple functional similarities with their more traditional counterparts, such as naval mines and torpedoes, enable us to shed some light on the possible legal requirements for sea drones. To date, no military manual has addressed these issues explicitly. Given the functional parallels with naval mines and torpedoes, States should be encouraged to reflect these analogies in future doctrinal guidance, thereby contributing to a clearer and more coherent legal framework for sea drones.
Legal relevance of sea drone status
As explained in the previous sections, sea drones could potentially be accommodated within the existing regulatory framework under one of two mutually exclusive categories: warships, as defined in UNCLOS Article 29, or means of warfare, as articulated in AP I Article 35.
The classification of sea drones extends beyond a purely theoretical exercise and has practical implications for their rights and obligations during both peace and wartime – in other words, under the law of the sea and the law of armed conflict. The following sections examine the key legal implications of each classification.
Sea drones in times of peace
States may use sea drones, similar to warships and other State vessels, for various purposes, including peacetime operations (e.g., border control, protecting communication lines, and anti-piracy patrolsFootnote 85). Thus, the status of sea drones as either warships or naval weapons generates concrete consequences during their operation, not only in war but also in peacetime.Footnote 86 There are three primary issues associated with sea drones’ status: firstly, their sovereign immunity; secondly, their navigational entitlements; and thirdly, the consequences of the use of force against them, under Article 51 of the Charter of the United Nations (UN Charter).
Assuming that a sea drone could be classified as a warship (setting aside, for the moment, its non-compliance with the requirements of UNCLOS Article 29), it would enjoy sovereign immunity, a status granted to vessels operated by the government for non-commercial purposes.Footnote 87 UNCLOS Article 32, included in the part of the Convention on territorial seas, states that “nothing in this Convention affects the immunities of warships and other government ships operated for non-commercial purposes [except the provisions included in subsection A and in Articles 30 and 31]”.Footnote 88 Then, UNCLOS Articles 95–96 underline that the sovereign immunity of warships also applies on the high seas and in exclusive economic zones (EEZs). Conversely, if a sea drone is considered a means of naval warfare, it would not ordinarily enjoy sovereign immunity privilege based exclusively on its classification. This is because its unmanned nature brings into question its legal status as an “extension of the State’s sovereignty”, without which a sea drone may be viewed as just an armed hull, as argued in the section above. On the other hand, if a sea drone operates as part of the State’s armed forces or is considered a government vessel used for non-commercial purposes, it might benefit from sovereign immunity due to its functional link, rather than its classification.Footnote 89
Regarding the navigational entitlements resulting from a sea drone’s possible status as a warship, the right of innocent passage requires analysis. The long-standing innocent passage right through a coastal State’s territorial sea, codified in UNCLOS Article 17, is granted to “ships of all States, whether coastal or land-locked”. It encompasses two scenarios: (a) traversing the territorial sea of a coastal State without entering its internal waters, or calling at a roadstead or port facilities located outside them; and (b) travelling to or from such waters or facilities.Footnote 90 Even though the warship’s right of innocent passage is not uniformly interpreted either in the scholarship or in State practice,Footnote 91 UNCLOS does not exclude the application of this provision to warships.
Thus, as a general rule, a sea drone enjoying warship status could exercise its right of innocent passage through a coastal State’s territorial seas as long as the coastal State does not consider such a passage as prejudicial to peace, good order or its security.Footnote 92 Also, UNCLOS Article 20 requires “submarines and other underwater vehicles … to navigate on the surface and to show their flag” during their transit via the territorial sea of a coastal State. Given that certain sea drones may operate underwater, this provision imposes additional peacetime constraints on their deployment.
In contrast, if classified as a weapon, a sea drone would lack any navigational rights, including the right to innocent passage.Footnote 93 This classification would also strengthen a coastal State’s ability to invoke the prejudicial nature of the sea drone’s presence within its territorial seas.Footnote 94 Moreover, such a situation – a sea drone’s presence within another State’s territorial sea – could be considered a violation of the coastal State’s sovereignty, or in certain cases, even an unlawful use of force.Footnote 95 Consequently, the lack of navigational rights and the other above-mentioned problems would significantly limit a sea drone’s operational area.
In the case of use of force against a sea drone, it is noteworthy that warships are considered “a political and military instrumentality of the State”.Footnote 96 This means that an act of violence against a sea drone that is classified as a warship could, depending on the circumstances, amount to an armed attack in the sense of the of UN Charter Article 51, and not merely an unlawful use of force.Footnote 97 Consequently, the flag State could respond within its right to self-defence.Footnote 98 On the other hand, if a sea drone is considered within the framework of means of warfare, a use of force against it would not, in itself, amount to an armed attack as discussed above, although it may still constitute an unlawful use of force under UN Charter Article 2(4).Footnote 99
Sea drones in times of war
Warships are the only types of vessels that enjoy belligerent rights under the law of armed conflict;Footnote 100 thus, the sea drone’s classification as such would allow it to engage in belligerent acts during an international armed conflict and to employ the means and methods of naval warfareFootnote 101 (e.g., enforcing a naval blockade or exclusion zoneFootnote 102). If they could be considered as warships, sea drones would also have the right not only to attack military objectives but also to visit, search, divert and capture enemy and neutral merchant vessels as per the rules on these topics contained in the law of naval warfare.Footnote 103 Moreover, as a matter of legal entitlements, they could undertake inspection of specially protected enemy vessels (e.g., hospital ships), as well as control neutral vessels and aircraft within operational zones. However, the practical exercise of certain rights, such as visit, search and inspection, presupposes human presence and judgement;Footnote 104 therefore, fully unmanned sea drones are likely unable to perform these functions in practice, even if they are classified as warships. Sea drones enjoying warship status would be entitled to demand that enemy personnel surrender and to undertake convoy operations,Footnote 105 and at the same time would be lawful military objectives by nature. For the reasons explained above, some States find it strategically advantageous to follow this classification.Footnote 106
Conversely, a sea drone considered as a naval means of warfare would lack its own belligerent rights. The State would enjoy belligerent rights, allowing it to deploy a sea drone just as with any other type of weapon such as a naval mine or torpedo,Footnote 107 the use of which would remain subject to the general restrictions on means and methods of warfare under IHL.Footnote 108 A sea drone would be simply a material asset of a State, rather than an entity capable of possessing its own rights or privileges.
Conclusion
Drones have been revolutionizing the commercial and military industries for a number of years, and now, sea drones are also transforming naval warfare. However, even though sea drones appear to be a highly innovative product of the twenty-first century, they share multiple legal characteristics with conventional military platforms and means of naval warfare, as demonstrated above. If manned, even if only with the crew operating the guns, they can be classified as warships under UNCLOS Article 29. On the other hand, a sea drone, while operating in an unmanned capacity independent from human involvement in its operation (human-in, -on or -off-the-loop), may constitute a means of warfare under IHL. The physical presence of the on-board crew constitutes a decisive factor in the mentioned classifications, resulting in various legal consequences, as explained in this paper.
It is likely that maritime powers, following the US example, will support a broader and more comprehensive warship status for sea drones,Footnote 109 granting them a full package of rights and privileges, as presented above. However, it is doubtful whether such a legal classification, based on a policy decision rather than on the strict application of the law, is desirable in light of the multiple legal and ethical questions related to the nature of autonomous and unmanned systems. As the discussion on the Newport Manual demonstrates, operational realities may inform classification debates, but elevating them above treaty-based criteria risks creating inconsistency and enabling States to classify unmanned systems at will.Footnote 110 In turn, it will likely undermine legal certainty and blur the distinction between vessels and weapons.
This dynamic underscores the need for the international community to engage more deliberately in shaping the legal framework, rather than leaving its evolution solely to the major naval powers whose practice has historically carried disproportionate weight. Accordingly, it would be important for other States to articulate their opinio juris on the status of sea drones, so that the development of the law does not default to an expansive interpretation driven primarily by the interests of powerful navies.
Finally, the humanitarian aspects of sea drones’ operation and compliance with the fundamental principles governing armed conflicts (e.g., distinction, precaution, proportionality and military necessity), not only at sea but in all domains, should guide their regulation and use. It is noteworthy that some possible regulatory solutions for sea drones already exist – given their resemblance to traditional means of naval warfare such as naval mines and torpedoes, rules governing the latter may be relevant. In any event, with sea drones’ continued development, the law of naval warfare must likewise adapt to safeguard the humanitarian principles that protect life in armed conflicts.
Open access
