1. Introduction
Autonomous and uncrewed vessels have begun to revolutionise the maritime landscape with their broad civil and military applications. The study of their impact is advancing but still limited, particularly in relation to certain types of these vessels. For example, while there are multiple research initiatives involving civil autonomous and uncrewed vessels around the world,Footnote 1 ongoing military projects using these new categories of vessels remain less known.Footnote 2 The main maritime navigation regulatory body, the International Maritime Organization (IMO), has focused on the new regulatory framework for merchant shipping autonomous and uncrewed vessels, excluding passenger vessels.Footnote 3 Academic literature notes that the existing legal framework is inadequate to govern the operation of these vessels, and has attempted to identify regulatory and legal barriers faced by autonomous ships when operating globally.Footnote 4
Although there is a growing body of literature on the lacunae and obstacles for autonomous shipping under the existing legal framework, much less attention has been devoted to theoretical problems posed by the law enforcement aspects, particularly under the extant regulations set by the United Nations Convention on the Law of the Sea (UNCLOS).
Technological advancements brought about by autonomous ships constitute unique challenges to law enforcement, broadly understood. Regulations pertaining to autonomous ships are scarce, and the existing legal framework has not been adapted to accommodate these ships, which could be viewed as ungoverned according to the classic triad ship-captain-crew concept.Footnote 5 Even accounting for the current regulatory deficit, ordinary enforcement by domestic and international agencies, including the attribution of responsibility, continues to be a clear challenge for autonomous ships.
This article seeks to position law enforcement as an essential consideration in the future regulation of civil autonomous shipping under UNCLOS, recognising the pressing need to accommodate this new category of vessels within the regulatory framework.Footnote 6 The article claims that international regulation must adapt to the challenges emerging from rapid technological advancements. In order to do so, regulations must be drafted bearing in mind that traditional forms of maritime law enforcement will not prove effective, and alternative systems to enforce regulations must be conceived of that are designed to interact with vessels that do not share many classic characteristics.
The contribution of this article to the literature is twofold. First, it focuses on law enforcement in autonomous shipping and identifies the challenges posed by these vessels to traditional maritime law enforcement. Second, it considers law enforcement regarding autonomous ships in relation to coastal and flag States’ enforcement powers.Footnote 7 In light of this, the article calls for further research on alternative forms of law enforcement to address the technological challenges posed by these vessels.
The remainder of the article is organised as follows. Section 2 revisits the concept of law enforcement by drawing from legal and economic literature. It argues that maritime enforcement has long presumed physicality—i.e. the physical presence of seafarers—as an element of enforcement. Physicality is absent in autonomous and uncrewed vessels where decision-making and decision-execution are carried out remotely or by the vessel itself. Section 3 tackles the enforcement challenges from the perspective of the key State actors under UNCLOS, focusing on coastal State enforcement powers in relation to maritime autonomous surface ships (MASS), innocent passage and UNCLOS provisions on criminal and civil jurisdiction, as well as pollution prevention in the context of MASS. Section 3.3 analyses the MASS-related enforcement powers of a flag State. Section 4 concludes by arguing that alternative forms of law enforcement ought to be explored as future regulation will be unable to rely on traditional forms of law enforcement effectively.
2. Law enforcement and autonomous shipping: a developing field
The concept of law enforcement is inherently linked to legal theory, sociology and political economy. Weber clearly defined a State in terms of its ability to enforce the law through legitimate physical force.Footnote 8 In turn, Kelsen’s well known account of law relied heavily on the idea of sanctions.Footnote 9 Weber and Kelsen are obviously far from being the first to concern themselves with law enforcement. Earlier scholars, from MontesquieuFootnote 10 to BeccariaFootnote 11 and Bentham,Footnote 12 considered several elements of law enforcement to be central to their theories of law, the separation of powers and criminal policy. Modern legal studies of law enforcement focus on the challenges of applying and executing regulations. At the same time, economics literature delves into deterrence theory and what resources are financially justified for apprehending violators.Footnote 13
2.1. The elements of law enforcement
Law enforcement is a central element of both national and international law, involving the use of governmental agents and/or international administrative agencies to detect and sanction those violating legal rules.Footnote 14 Law enforcement attempts to ensure the effective functioning of a legal system. In general, it refers to creating legal regulations intended to induce compliance, monitoring adherence to legal rules and implementing sanctions imposed to deter violations.Footnote 15
In international law, law enforcement is not straightforward due to certain limits on physical coercion as an enforcement mechanism. Even if States Parties to international treaties and international bodies draft enforcement-related regulations, monitor the application of legal rules and have formal and informal sanctions to induce compliance and deter violations, the nature of international law enforcement is still broadly debated.
Austin famously questioned whether international law was law, partially due to the lack of an arbiter to enforce rules,Footnote 16 an objection rebutted by Hart.Footnote 17 Subsequently, scholars have noted that enforcement of international law does not necessarily match traditional notions of law enforcement insofar as the latter requires possessing internal enforcement mechanisms and the possibility of employing physical force to induce compliance.Footnote 18
It is generally considered that international law has relied on the concept of territory to delimit jurisdiction. The predominance of territorial concerns in defining jurisdiction is a classical part of the Westphalian conception of the international legal order.Footnote 19 In the so-called first principle of the Lotus case, the Permanent Court of International Justice (PCIJ) held that international law imposes a restriction on States: ‘jurisdiction is certainly territorial; it cannot be exercised by a State outside its territory except by virtue of a permissive rule derived from international custom or from a convention’.Footnote 20 Hence, absent a permissive basis, States were considered unable to exercise extraterritorial enforcement jurisdiction. Although contested, this doctrine has undoubtedly impacted the International Court of Justice’s jurisprudence and remains a part of the international law canon.Footnote 21
While international law scholars have thoroughly examined notions of enforcement, the multifaceted complexities presented by autonomous shipping are particularly challenging for the enforcement of law of the sea regulations, which the next section will illustrate.
2.2. Law enforcement in relation to autonomous shipping
Law enforcement is a crucial concept in both legal scholarship and practical application. Notwithstanding that a significant part of the literature focuses on ordinary and criminal law enforcement,Footnote 22 law enforcement at sea has emerged as a distinct field warranting systematic study beyond isolated issues. Jimenez analysed the distinction between maritime law enforcement and the use of force at sea.Footnote 23 Soons showed that law enforcement at sea is sectorised through several jurisdictional layers corresponding to maritime zones,Footnote 24 reaching conclusions similar to those of Klein.Footnote 25 Sun et al delved into technical obstacles regarding maritime intelligence, which plays a crucial role in higher-quality law enforcement in maritime domains,Footnote 26 while Jiang and HuangFootnote 27 analysed the element of territoriality within maritime enforcement in a case study of China’s Marine Environmental Protection Law, following its revision in 2023. While some legal scholars have begun to examine the impact of autonomous shipping on the international legal order,Footnote 28 most concentrate on formulating policy rather than on the key issues of monitoring, sanctioning and deterring non-compliance.
Since legal scholars and practitioners define autonomous and uncrewed ships differently,Footnote 29 a shared typology is needed to enable clear analysis across maritime zones. In 2018, the IMO introduced the working term ‘Maritime Autonomous Surface Ship’ (MASS) in the Maritime Safety Committee’s Regulatory Scoping Exercise.Footnote 30 Since then, IMO’s MASS classification has entered the legal jargon alongside four proposed degrees of autonomy, namely: Degree I—a ship with automated processes and decision support; Degree II—a remotely controlled ship with seafarers on board; Degree III—a remotely controlled ship without seafarers on board; and Degree IV—a fully autonomous ship.Footnote 31 Even though the IMO MASS classification ignores nuanced design variations and legal complexities arising from different levels of ship autonomy and crewing,Footnote 32 it is sufficient for examining the law enforcement mechanisms.
Nevertheless, it is worth noting that terms such as ‘autonomy’ and ‘(un)crewing’ are not synonyms, even though they are often employed interchangeably. Autonomy refers to the operational decision-making process, while (un)crewing simply signals onboard human presence, or the lack thereof.Footnote 33 Consequently, there might be various combinations of both factors—autonomy and crewing—leading to multiple categories of vessels enjoying different levels of operational autonomy with varying numbers of onboard crewmembers. Additionally, their categorisation is complicated by the fact that vessels might only be autonomous and uncrewed for a particular voyage or part of one.Footnote 34 Finally, a shore-based Remote Operations Centre (ROC) together with a remote operator (RO) are always presumed in the discussion of autonomous and uncrewed vessels as the state of technology still does not allow the operation of fully autonomous and uncrewed vessels without at least remote human engagement or supervision. Section 3 examines challenges to law enforcement regarding autonomous shipping vis-à-vis key State actors’ enforcement powers, as provided for by UNCLOS.
3. Key State actors’ enforcement powers under UNCLOS
Enforcement, as a concept, presumes an enforcer. In international law, States enforce international regulations to different degrees and under different conditions. There are several forms in which a State is represented as an entity in UNCLOS.Footnote 35 The primary State actors are coastal and flag States,Footnote 36 which are central to the UNCLOS enforcement framework in terms of international shipping.Footnote 37 Their roles reflect the varying degrees of State obligations, State control and jurisdictional complexities, although they safeguard the foundational balance of interests between them.
To add further complexity, seas are divided into different maritime zones and areas, described by two main characteristics: their breadth and the State rights exercised therein.Footnote 38 The contemporary construction of the maritime zones is derived from UNCLOS, which sets the governing regulations for each zone and assigns different rights to the interested States. Treves calls them ‘exclusivity rights’, that is, rights to undertake certain activities to the exclusion of other States.Footnote 39
Scholars classify maritime zones in different waysFootnote 40 but, for the purposes of this article, three main categories can be discerned, reflecting the varying degrees of State control and jurisdictional complexity: first, sea areas under the sovereignty of a coastal State; second, sea areas under the functional jurisdiction of a coastal State; and, third, sea areas beyond the jurisdiction of coastal States. These classifications—of both the State actors as well as the maritime zones—have significant implications for law enforcement as they impact the regulations to enforce, the parties responsible for and authorised to enforce them and even the available legal and de facto enforcement measures. In other words, pairing a State actor with a specific maritime zone clarifies who can do what in that area of the sea.
The coastal State enjoys sovereignty over internal waters, the territorial sea and the archipelagic waters.Footnote 41 This gives the coastal State broad enforcement powers over these maritime zones, with certain limitations based on the right of innocent passage enjoyed by foreign vessels, which is further considered in Section 3.1.
In the sea areas under the functional jurisdiction of a coastal State—which include the contiguous zone, exclusive economic zone (EEZ) and continental shelfFootnote 42—the coastal State enjoys limited enforcement powers, especially in relation to customs, fiscal, immigration and sanitary laws (contiguous zone)Footnote 43 and to living and non-living resource management (EEZFootnote 44 and continental shelf).Footnote 45 As all the zones under the functional jurisdiction of a coastal State lie outside the outer limit of the territorial sea of a coastal State, as a general rule, they are governed by the freedom of navigation.Footnote 46 Consequently, vessels remain under the exclusive jurisdiction of a flag State, which is analysed in Section 3.3, and the coastal State may only exercise enforcement measures over vessels violating the laws established for each zone in a few specific cases.Footnote 47 However, those cases typically concern activities such as fishing, pollution control or resource exploitation, which are not currently the operational focus of MASS as used in international shipping.
Consequently, the enforcement challenges related to MASS navigation are similar to those occurring within the territorial sea of a coastal State, discussed in detail in the next section. In terms of the contiguous zone and EEZ, enforcement measures are limited to specific activities, but the practical problems remain the same.
3.1. Coastal State enforcement powers against MASS in relation to innocent passage
In the coastal State’s territorial sea and archipelagic waters, as a general rule, a foreign vessel enjoys the right of innocent passage.Footnote 48 The concept of innocent passage has been broadly analysed in the literature due to its well-established status as a rule of customary international law.Footnote 49 For present purposes, it suffices to note that it is a right composed of two parts, ‘innocence’ and ‘passage’, to traverse the territorial sea of a coastal State without entering internal waters or in order to proceed to or from them.Footnote 50
The innocence requirement is met as long as the vessel does not engage in activities that might be prejudicial to the peace, good order or security of the coastal State.Footnote 51 However, the coastal State has the right to undertake all necessary steps to prevent passage through its territorial sea if a vessel violates this rule.Footnote 52 In other words, the coastal State enjoys full enforcement powers over a foreign vessel if the passage does not meet the innocence requirement.Footnote 53 It may stop, inspect and divert from its territorial sea any vessel engaging in violating conduct.Footnote 54 The coastal State may also detain and force such a vessel to its port in order to undertake corresponding legal proceedings.Footnote 55 Although these measures are primarily aimed at the vessel itself, they usually negatively impact the crew members onboard by subjecting them de facto to the coastal State’s enforcement powers. This may result in additional stress, inconvenience, delays in crew exchanges and destabilisation of seafarers’ working and living conditions, to name just a few impacts.
However, not every activity violating the coastal State’s regulations automatically means a violation of innocent passage (e.g. non-compliance by a vessel with a local traffic separation scheme established to enhance navigational safety usually does not qualify as prejudicial under Article 19(2) UNCLOS). In such cases, the coastal State’s enforcement powers are more constrained (e.g. limited to requiring the vessel to comply with the regulation, but not ordering the ship to leave the territorial sea).Footnote 56 Therefore the violations of the coastal State’s regulations must always be examined to verify whether they fall under the scope of the non-innocent behaviours established by Article 19 UNCLOS. As usual, UNCLOS offers only general guidance regarding the coastal State’s enforcement measures in terms of the territorial sea and innocent passage. This reflects the fact that the territorial sea is subject to the full sovereignty of the coastal State, whereas in the EEZ, the coastal State has only sovereign rights and broader navigational freedoms apply. More concrete approaches are derived from State practice rather than from the straight normative formulation of UNCLOS.Footnote 57
MASS, due to their operational autonomy and uncrewed nature, are both the source of conceptual difficulties in defining what constitutes innocent passage and the object of the practical difficulties in enforcing coastal State regulations related to it. The situation of MASS Degrees I and II is much like that of conventional vessels since having a crew onboard aids compliance with a coastal State’s enforcement actions if the passage threatens its peace, good order or security. However, coastal State enforcement actions against MASS Degrees III and IV will encounter compliance problems without humans on board who can react to the coastal State’s requests to stop, undergo an inspection or divert from the territorial sea. In the case of conventional vessels, coastal States can exercise jurisdiction regarding innocent passage over both the vessel and the crew, including the captain and the officers in charge. In the case of MASS Degrees III and IV, jurisdiction would be exercised over the vessel and attempted to be exercised over personnel located in the shore-based ROC, possibly in a different jurisdiction. The type of interaction between the coastal State authority and the ROC would vary greatly depending on the alleged non-innocent behaviour. Actions may range from simple instructions to cooperate with coastal State authorities in arresting the vessel pending the posting of a reasonable bond or otherwise risk vessel damage during its arrest, to more significant exercises of jurisdiction, such as arrest and prosecution.Footnote 58
Detention procedures, especially in the coastal State’s port, may be the most challenging enforcement measure in the case of an innocent passage violation by a MASS of Degrees II, III and IV, as any action will require remote coordination with the respective ROC. MASS Degrees II–IV offer a paradoxical example in which the consequence of human behaviour (i.e. the movement of the vessel) is geographically within the enforcement area of a specific State, but the human commanding or responsible for the behaviour is not.
The traditionally understood enforcement measures (e.g. stopping, inspecting, diverting and detaining) related to the violation of innocent passage by foreign vessels are meant, in practice, to target both the vessel and its crew. The lack of physical presence of the captain and crew onboard shifts the compliance responsibilities to the remote operators based in other parts of the world. In the case of MASS Degree IV (fully autonomous and uncrewed), operated by artificial intelligence (AI), the enforcement mechanisms would need to target AI, which per se cannot be sanctioned by physical detention in a foreign port, for example.Footnote 59 Legal scholarship is just beginning to explore innovative liability systems in relation to AI,Footnote 60 such as proportional liability, which should be considered in instances involving MASS Degree IV.
3.2. Coastal State enforcement powers against MASS in the UNCLOS special provisions on criminal and civil jurisdiction, as well as environmental pollution
The other category of enforcement mechanisms regarding foreign ships engaged in innocent passage arises from the specific provisions governing criminal and civil jurisdiction in relation to those vessels,Footnote 61 which allow the coastal State to exercise a degree of authority over them.Footnote 62
The distinction between a coastal State’s jurisdiction over a vessel or a person is crucial. According to Article 27 UNCLOS, the person suspected of the crime is subject to the enforcement measures, rather than the vessel where the crime was committed.Footnote 63 In the case of MASS Degrees I and II, the application of criminal jurisdiction provisions would apply in the same manner as with conventional vessels due to the physical presence of the crew onboard. However, for MASS Degrees III and IV, their uncrewed nature would complicate direct application of criminal jurisdiction. The engagement of the remote operator and other shore-based personnel would require the reevaluation of criminal jurisdiction over ship-related personnel physically based in a distant location and not onsite where the consequences of the criminal act occurred.Footnote 64 Perhaps in this case policymakers should consider, de lege ferenda, drafting international agreements with innovative jurisdictional frameworks, such as treating shore-based personnel as if they were physically present in the territorial sea of a coastal State, including extradition agreements and an inter-State assistance obligation. Despite measures of this kind, in practice the execution of the enforcement measures vis-à-vis shore personnel based in a remote location may still be difficult and often de facto impossible.
The situation is reversed in the case of civil jurisdiction, which usually relates to contractual and non-contractual obligations, labour relationships and property regulations.Footnote 65 As a general rule, the coastal State should not exercise enforcement measures such as the detention of a person or a vessel for the purpose of civil proceedings. However, it may indeed levy execution against or arrest a vessel regarding the ‘obligations or liabilities assumed or incurred by the ship itself in the course or for the purpose of the voyage through the waters of the Coastal State’.Footnote 66
These provisions on civil obligations and liabilities assumed by the vessel during its voyage presume that a real person is responsible for the decision-making process. Thus, MASS Degrees I and II could still be subject to enforcement measures by the coastal State as there would be humans engaged in and responsible for the decision-making process. The situation regarding autonomous and uncrewed vessels is less straightforward. MASS Degrees III and IV operate based on pre-programmed algorithms, with decision-making reliant on AI. Consequently, the attribution of civil liability to autonomous vessels would require the assessment of the roles played by other actors, such as software developers and manufacturers,Footnote 67 or reevaluating liability theories, as has recently been proposed.Footnote 68
However, this limitation on civil jurisdiction does not prevent the coastal State from levying execution against or arresting a foreign vessel in relation to civil proceedings when the vessel is present in its territorial sea or navigating the territorial sea after leaving the coastal State’s internal waters.Footnote 69 In this scenario, the physical presence of the crew onboard MASS Degrees I and II, as well as the involvement of the remote operator—MASS Degree III—could facilitate easier compliance with the coastal State’s requests. However, the coordination of activities with the distant ROC would still impair the full process. The coastal State’s execution of arrest warrants and further enforcement actions regarding MASS Degree IV would be even more complicated given its fully autonomous and uncrewed nature. Additionally, as mentioned before, the enforcement of both civil and criminal jurisdiction vis-à-vis vessels traditionally involves human interactions and onboard inspections. Thus, the shore-based ROCs, as well as the AI in charge of MASS Degree IV, would also need to be technologically prepared for such possibilities.
Aside from the general rules applicable to the violation of innocent passage, special UNCLOS regulations allow coastal States to take measures if a vessel navigating its territorial sea or EEZ may have violated rules on the prevention, reduction and control of pollution from vessels.Footnote 70 In this case, the coastal State may initiate a physical inspection of a vessel suspected of the violation and is entitled to institute proceedings based on the supporting evidence.Footnote 71 The vessel’s detention is a permitted enforcement action.Footnote 72 In the case of MASS Degrees I and II, the physical presence of crew onboard, even without a bridge crew, would still allow the crewmembers to undertake immediate corrective actions to prevent further harm to the marine environment. However, in the case of MASS Degrees III and IV, the immediate mitigation of environmental pollution and compliance with enforcement requests from the coastal State may be especially challenging, as it usually requires physical action onboard.Footnote 73
3.3. Flag State enforcement powers in the context of MASS
According to longstanding maritime custom—codified in Article 92 UNCLOS—vessels, as a general rule,Footnote 74 are subject to exclusive flag State jurisdiction on the high seas, a key principle of the law of the sea.Footnote 75 The freedom of navigation governs this area beyond the national jurisdiction of States,Footnote 76 which remains open to all of them.Footnote 77 Article 91 UNCLOS on vessel nationality requires the existence of a ‘genuine link’ between the State and the vessel flying its flag. As UNCLOS does not specify the meaning of genuine link, the interpretation of this requirement has been the source of legal dispute.Footnote 78
The International Tribunal for the Law of the Sea (ITLOS) stated in M/V “Saiga” (No 2) that ‘the purpose of the provisions of the Convention on the need for a genuine link between a ship and its Flag State is to secure more effective implementation of the duties of the Flag State’.Footnote 79 Moreover, the flag State is obligated to exercise an effective ‘jurisdiction and control in administrative, technical and social matters’,Footnote 80 which ITLOS also confirmed in M/V “Virginia G”, where it underscored that ‘once a ship is registered, the Flag State is required, under article 94 of the Convention, to exercise effective jurisdiction and control over that ship in order to ensure that it operates in accordance with generally accepted international regulations, procedures and practices. This is the meaning of “genuine link”’.Footnote 81
In the case of MASS Degrees I–IV, determining a genuine link between the vessel and the flag State is even more complex than on conventional vessels. In the latter case, the vessel’s registry—and, thus, its nationality—is often chosen taking into consideration factors such as the operational costs, labour regulations or even the characteristics of the State where the ship’s operator is domiciled.Footnote 82 Given that all categories of MASS usually involve shore-based ROC and remote operators, the locations of those additional actors, non-existent in conventional situations, would also need to be considered when establishing the genuine link and flagging the ship.Footnote 83 A shore-based ROC in a State that is not the flag State could disrupt traditional notions of jurisdiction vis-à-vis enforcement. New notions of law enforcement, such as theories granting jurisdiction to the State where the ROC is located or the State where the company owning the ROC is legally registered, should be considered.
Failure to address the critical challenges posed by remote locations of the ROC may disrupt the enforcement of flag State jurisdiction and the current version of the genuine link doctrine might be insufficient to address the new challenges posed. As mentioned in Section 2.1, territorial concerns have greatly shaped the definition of jurisdiction in international law and, to a certain extent, in the law of the sea. Territorial definitions of jurisdiction have obvious advantages: they are based on the physical location of behaviour and appeal to notions of State sovereignty. However, territoriality’s centrality arose in a technologically different era in which the geographical location of behaviour was clear, which it may not be in relation to MASS which comprise both the vessel and the operational centre. Although these considerations are being raised in relation to flag State enforcement powers, given their significance in light of the genuine link doctrine, similar considerations are applicable, mutatis mutandis, also in relation to coastal State enforcement powers, which also depend on identifying a territorial nexus to assert jurisdiction.Footnote 84
How should international law tackle the regulation of emerging technologies where behaviour might be cross-territorial, with command and execution taking place in different locations that might be continually changing? One possibility would be to develop joint jurisdiction doctrines, such as dynamic or fluid jurisdiction, in which regulations would assess jurisdiction based on which State is best placed to effectively enforce the law in practice, rather than the characteristics of the vessels. Future international regulations might recognise the difficulties of solely relying on the territorial location of behaviour and conceive of multiple States as simultaneously competent to enforce certain regulations; or the regulations themselves may be considered dynamic, hence defining enforcement capability by the suitability of the enforcer rather than by the territorial location of the controlled behaviour. While it is acknowledged that such characteristics might not directly correspond to a traditional notion of jurisdiction as shaped by territorial concerns, they might be necessary to ensure that rules are enforceable and that MASS can be comprehensively regulated. Under current law, the involvement of a shore-based ROC located in a place other than the flag State could lead to more than one State claiming jurisdiction over the vessel. For example, the State where the ROC is based could invoke territorial jurisdiction since the activities governing the MASS operation occurred within its borders. Similarly, cybersecurity and data protection regulations and mechanisms could result in jurisdictional overlap with the flag State due to MASS’ reliance on data transmission, indeed, such overlap would be a common occurrence in MASS operations. Currently, some national legislation contemplates certain situations in which jurisdiction is exercised according to a variety of considerations, such as the location of the vessel, despite the lack of a specific classic jurisdictional premise (such as a clear single location for the act). Examples are arguably present in Australia’s Crimes at Sea Act, the Spanish Law on Maritime Navigation and Article 110 of the Polish Penal Code.Footnote 85
It is also doubtful whether current technology supported by legal mechanisms allows the flag State to exercise effective jurisdiction over MASS Degrees III and IV, being unmanned and operated fully or to a certain degree by AI. Real-time monitoring, remote and AI decision-making, as well as enforcement of compliance depend on uninterrupted and smooth data transmission and secure communications necessary for the safe and secure navigation of these new categories of vessels.
The complexity of the genuine link determination becomes especially prominent when the phenomenon of flags of convenience (FoC) is considered. This is where a vessel has no national connection to the registry, but has been attracted by, for example, the ease of the registration process, lower taxes, less stringent labour and environmental regulations or lower inspection standards. Such practices jeopardise maritime safety and security.Footnote 86 The emergence of new categories of autonomous and uncrewed vessels will likely lead to the advent of an ‘autonomous flag of convenience’, which could foster sub-standard and dangerous autonomous vessels. Having no physical human presence onboard will impede the effective execution of enforcement mechanisms by interested States to an even greater degree.
Finally, an analysis of enforcement in relation to MASS must also consider the transit passage enjoyed by vessels through the straits used for international navigation.Footnote 87 Even though, at first glance, the right of transit passage looks similar to innocent passage—as it may also be exercised over the straits used for international navigation and archipelagic sea lanesFootnote 88—the coastal and flag State positions are reversed.Footnote 89 Unlike with innocent passage where the coastal State may take measures against a ship which is not innocent, as a general rule, the coastal State may not impede transit passage.Footnote 90 In practice, this results in primary flag State jurisdiction and restricts the execution of enforcement measures by coastal States such as inspections, detentions or arrests.Footnote 91 Consequently, the issues arising in relation to enforcement against MASS where the right of innocent passage is breached also arise with respect to transit passage.
4. Conclusions
Autonomous shipping poses a significant challenge to law enforcement at sea. Although the law of the sea has always divided law enforcement among the key State actors according to maritime zones, the additional jurisdictional and enforcement issues raised by automated or remote decision-making and the lack of a human crew require a re-conceptualisation of law enforcement to accommodate autonomous ships.
This article has examined the ambiguities that arise in relation to law enforcement, demonstrating the inadequacies of the current legal framework to govern autonomous shipping. Three key features of autonomous ships that present challenges for law enforcement have been identified: first, the lack of a fixed location to determine jurisdiction; second, the lack of a physical presence of crew and captain, complicating effective enforcement; and, third, the use of automated or remote decision-making, which undermines the central deterrence function of sanctions.
Further, it has been shown that, methodologically, enforcement by State actors should vary according to the different categories of MASS and the particular maritime zone in which they are located. By analysing the challenges posed by autonomous ships in different maritime zones, the article hopes to initiate a debate on the new regulatory tools required, particularly when assessing the compliance of autonomous shipping with UNCLOS. It suggests that new regulations must be drafted with alternative forms of enforcement that dispense with physicality. Similar challenges have been raised in other areas of law. Scholars analysing cyberspace conductFootnote 92 or space lawFootnote 93 have long abandoned traditional notions of law enforcement and have begun seeking new, more effective tools. In the case of cyberspace, the lack of physicality forces scholars and policymakers to conceptualise alternative notions and means of enforcement, while space law contends with enforcement agency access and technological barriers to enforcing compliance.
Among other actions, it is necessary to clarify the genuine link requirement in terms of these new categories of vessels and the enhanced flag State responsibilities they entail. Finally, the exclusivity of flag State jurisdiction over autonomous and uncrewed vessels on the high seas must be reevaluated, as it poses significant limits on enforcement in relation to MASS by the international community. It is also crucial to establish remote enforcement mechanisms.
Potential solutions were raised which may include devising international agreements that establish dynamic or fluid enforcement mechanisms, changing the traditional person/object/flag of registration view of jurisdiction to an enforcement based on convenience view. Modifying how jurisdiction is determined from the territorial location of behaviour to simultaneous jurisdiction or enforcement convenience represents abandoning longstanding canons of international law, but perhaps it is necessary to recognise that those rules are inadequate for ships that are controlled remotely or by AI.
While further analysis is required to find workable solutions, this article suggests that physicality as a central element of law enforcement should be reexamined in relation to such vessels. Autonomous shipping will require the creation of innovative legal doctrines or rethinking the prevailing ones. From the territorial waters to the high sea, it is high time.
Acknowledgements
This article presents the results of a research project entitled ‘Autonomous and Unmanned Vessels in International Law: Yara Birkeland as a Challenge to the Law of the Sea’ that has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 101038097. Mauro Arturo Rivera León’s research activities were co-financed by the funds granted under the Research Excellence Initiative of the University of Silesia in Katowice.