Abstract
The end of the Cold War era has ushered in a century in which great military powers like the US, China, and Russia continue to dominate the arms trade, yet minor powers are also investing heavily in arms manufacturing and technologies and profiting from lucrative arms transfers. Changes in the arms trade have fostered cooperative international regulatory practices and agreements, but recent conflicts in Yemen, Syria, Libya, Nagorno-Karabakh, and Ukraine highlight the need for strong end-use monitoring, post-shipment verification, and international accountability for dubious trade deals. This study analyzes changes in top arms producers, new complexities in arms regulation due to advancing technologies and artificial intelligence, arms-control policies of international organizations and top arms producers, and the risks of poor regulation.
Keywords
Introduction: Arms trade in the twenty-first century
During the late twentieth century, the international arms trade was dominated by the two reigning superpowers of the post-war bi-polar system, the US and Soviet Union, which together accounted for 70 percent of the world's arms transfers. 1 Though the US and Russia remain the top arms producers in the world today, and though some reports estimate that China has overtaken Russia through unreported sales, many regional powers and advanced industrialized states are becoming leaders in arms technology, production, and trade, thus changing the nature of the arms trade. Countries such as Canada and Sweden, long viewed as international peacemakers, have found themselves consistently ranked within the top fifteen arms sellers in the past decade. 2 Regional powers that historically relied on US or Russian military aid, such as Saudi Arabia and Pakistan, have found an ever-growing array of sellers. Rising regional powers with complex security motivations, such as Israel and South Korea, have invested heavily in advanced military technology and artificial intelligence (AI). States such as Iran, restricted from access to advanced technologies by Western sanctions, have begun producing their own components and become regional sellers in combat drones. 3 Non-state military groups have access to high-tech military armaments through gray and black markets, fueling the many civil wars that have erupted in the post-Cold War era. 4
The growth of private arms manufacturers and traders, the lucrative benefits of exporting arms, the growing availability of conventional weapons systems, the spread of information through cooperative or joint development, and improved dual-use and advanced military technologies have all created a dilemma for those concerned with international security and arms control. Thus, the focus of this analysis is in understanding contemporary trends in the international trade of arms, including shifting relationships between top exporters and importers, the rise of minor powers as arms-exporting leaders, international and state-level attempts at regulation, advancing technologies and the dual-use nature of military technology, and the risk of fueling conflict by prioritizing trade over-regulation enforcement.
Shifting arms supply patterns
The international trade and transfer of conventional arms, though retaining many carryover patterns, has evolved significantly in the past two decades. In 2002, with declining international wars, global arms imports sat at the lowest levels since 1963, totaling roughly $17.8 billion USD. By 2017, however, arms imports rose again to nearly $32 billion, their highest level since the collapse of the Soviet Union. 5 The US alone, as the long-standing leading global arms supplier, was responsible for $640 billion in arms sales between 2002 and 2019, primarily to allied or client states in the Middle East, such as Saudi Arabia, Israel, and the United Arab Emirates (UAE). 6
Arms manufacturing and exporting can also be a lucrative trade for advanced industrial states with smaller economies. Canada earned nearly $2 billion CAD from arms exports in 2020 (down from $3.75 billion in 2019), 75 percent of which were transfers to non-NATO countries. Of all Canadian arms transfers to countries other than the US, 67 percent were to Saudi Arabia—indicative of Ottawa's continued prioritization of a troubled trading relationship. 7 This trend exemplifies the lure of sales for wealthy nations, reflected in a $23 billion agreement for the US also to sell F-35 fighter jets to Saudi Arabia, despite reported shortcomings in those planes’ performance and ongoing end-use controversies involving Saudi Arabia and the UAE. 8 Canada completed an estimated $14 billion CAD deal to manufacture Light Armoured Vehicles (LAVs) for the Kingdom, despite calls for Ottawa to cease trading amid reported Saudi human rights abuses and end-use violations. 9 This trend also signifies the strategic shift from Cold War proxy warfare to attempted “stabilizing” missions in the Middle East and Africa by Western democracies motivated by theories of democratic peace, counter-terrorism, “nation-building,” and the containment of countries such as Iran. 10
Though the US remains the perennial world leader in arms transfers and exports, its share of the global trade was much diminished. During most of the post-war twentieth century, 90 percent of arms exports came from just five states: the US, Soviet Union, France, Germany, and the UK. 11 This is no longer the case, as China has risen to within the top-five leading arms exporters, although much of their sales are thought to go unreported. Chinese arms manufacturers accounted for at least $54 billion in reported arms sales in 2017, earning roughly $16 billion more than Russia. 12 Conversely, the UK fell to only 3.3 percent of global arms export sales between 2016 and 2020 (roughly equivalent to Israel), while France rose to the third rank with 8.2 percent of the global market. 13
Some attribute these changes to the increasing costs of purchasing technologically-complex American and Western arms and the increasing capability of minor producers to manufacture inexpensive conventional weapons. 14 However, rising security concerns in Europe and the Pacific may increase demand for expensive arms with the understanding that complex weapons may become a necessary defence against the security threat of regional powers. One such example is the controversial AUKUS nuclear submarine deal, which saw Canberra cancel an estimated $66 billion deal with the French Naval Group 15 after production delays and growing expenses for twelve diesel-electric powered attack-class submarines in favor of costlier, and possibly more delayed, nuclear submarines. 16 This example highlights rising regional tensions in the Pacific and the need for Canberra to compete with a superior Chinese navy with more complex naval capabilities. Though states may now seek cheaper arms alternatives (enabling the rise of new producers), contemporaneous security concerns may also bolster the production and sale of complex arms from the top manufacturers.
Another explanation for the diminished reliance of minor powers and peripheral states on traditional great-power producers are changing patterns in “hegemonic relations.” 17 With the collapse of the Soviet Union, and contemporary changes in great-power politics paired with the increasing availability of arms from minor producers, states use arms-transfer deals as political bargaining chips to foster security alliances and economic pacts. These politically-motivated arms deals may explain why American allies continue to enter into purchasing agreements for the F-35, despite ongoing production flaws and rising manufacturing costs 18 ; if a state relies on American trade or military security, it is more likely to buy American arms. 19 Similar concerns and bargaining strategies may also explain the relative decline in some arms producers, such as Russia, as states seek to distance themselves politically amid the war in Ukraine and to avoid potential production delays due to increased Western sanctions. 20 Additionally, such political bargaining may explain the sudden rise of China, as some states seek to circumvent Western arms embargoes or establish security alliances. One such example is Venezuela's reliance on Chinese arms, accounting for nearly 86 percent of all Chinese arms sales to the Latin American region. 21
Now among the top arms producers are states that are regarded in the international community as perpetual “humanitarian superpowers.” 22 Canada and Sweden, for example, have become top arms exporters in their own right, ranking sixteenth and fifteenth respectively in exports (measured in TIVs) between 2015 and 2021, 23 with some claiming that Canada ranks much higher if the production of component parts is taken into account. 24 The balance between maintaining its image as a peacemaker/peacekeeper, while benefiting economically from the transfer of arms, is a conflict Canada has been wrestling with internally, with many calling for Ottawa to investigate and cease its sales to countries with human rights violations, including Saudi Arabia and Israel. 25
The shifting relationships between leading supplier states and key importers are evidence of changing geopolitical security goals and the lure of economic gain from dubious deals. In 1988 the top arms-importing states by percentage of international share were Iraq (9.5%), India (6.6%), Saudi Arabia (6.2%), Afghanistan (5.3%), Iran (4.1%), and Israel (3.9%). 26 These top importers reflected the significance of the Middle East as a battleground for proxy warfare between the world powers and leading arms exporters; of course, subsequently, the US and its allies became enmeshed in several Middle Eastern and Central Asian conflicts in the “war on terror” and tried to prop up local militaries with armament. During this period of proxy warfare and alliance-building relationships, Iraq imported nearly $61 billion in arms, the majority from the Soviet Union, while the US sold $11 billion in weapons to Iran prior to the Iranian revolution, 27 with the glaring irony that it did not prevent the revolution.
With the end of the Cold War and the transition into the twenty-first century, strategies of remote or indirect warfare became preferred methods for military intervention, especially after costly wars in Iraq and Afghanistan. 28 Some states also wanted to keep their roles in conflicts covert, leading to a shift towards the acquisition of unmanned aerial vehicles (UAVs) and combat drones, requiring states to produce or obtain advanced electronic components. Additionally, global geopolitical security strategies shifted, as is indicated by the change in top arms importers. Between 2016 and 2020, the main customers of the US were Saudi Arabia (24%), Australia (9.4%), and South Korea (6.75%). 29 Moreover, the US transferred roughly $183 TIV (trade indicator value) in arms to NATO allies between 2018 and 2020. 30 Additionally, during the past decade Canada's top arms-trading partners, other than the US, were Saudi Arabia (307 million in TIV) and the UAE (145 million in TIV). 31 Despite human rights and end-use concerns, especially regarding the use of Canadian LAVs in Yemen, 32 and despite the irony of a self-declared feminist prime minister increasing arms sales to a Saudi regime notorious at the time for suppressing women's rights, 33 Ottawa continued to supply these states as its top arms-trading partners.
Arms sales to the Saudis remain a lucrative business for the US and Canada, and although concerns have mounted as to their end use, especially in the highly destructive Yemen civil wars, both countries have increased their weapon sales efforts. Despite a temporary halt on Saudi arms dealings pending investigation of their use and possible human rights abuses, both at odds with US arms-transfer regulations, the Biden administration approved a $23 billion deal with the Kingdom, including F-35 fighter jets and armed Reaper drones, 34 doubling the average annual US sales between 2015 and 2020. While Canada suspended any issue of new export permits to Saudi Arabia between October 2018 and April 2020, with a consequent decline in overall 2020 sales, it has since resumed issuing permits on a case-by-case basis. 35
Unlike the Canadian arms trade, which seems to prioritize economic strategy, the development of American arms exports to Australia and South Korea indicates a shifting security strategy, emphasizing a “pivot” of American security concerns from the Middle East to China, East Asia, and the Indo-Pacific. By bolstering South Korean and Australian military capabilities, the US seeks to implement a strategy of off-shore balancing to check the rise of China and secondarily, North Korea. 36
Alternatively, Russian's invasion of Ukraine has restored a degree of strategic restraint towards Russian expansion in Europe, as made evident by the nearly $10 billion in security assistance that the US initially committed to the defence of Ukraine, including the transfer of High-Mobility Artillery Rocket Systems (HIMARS) and artillery munitions, 120 mm mortar systems, surface-to-air missile systems (NASAMS), Javelin and AT-4 missile launchers, and a variety of demolition munitions including C-4 and claymore explosives. 37
The role of non-state actors and unconventional markets
As civil wars displaced international fighting in the early post-Cold War years, weapons production and distribution through failed or failing states (e.g., Libya), second-hand sales, private and black-market dealers, and major and middle power supplies all increased the military capabilities of non-state actors. 38 For example, Canadian small arms that were originally transferred to Kurdish forces engaged in anti-ISIL operations have been reportedly sold via black-market networks in Iraq, and Canadian sniper rifles have been both transferred deliberately to, and seized by, forces fighting in Yemen. 39
In a more explicit display of the variety of arms available through unconventional markets, the Libyan National Army militia faction, led by Field Marshal Khalifa Haftar, has acquired a variety of arms, including fifty surface-to-air missiles, four Mi-24P combat helicopters, and five MiG-21MF fighter jets from their Russian and UAE patrons. 40 These arms were reportedly all accessed through “gray market” sales, that is, covert transfers made by governments to avoid international restrictions or embargoes. 41 This occurred even as Libya was formally placed under a UN arms embargo, an embargo that a UN panel declared “totally ineffective.” 42 Apart from fueling the Libyan civil war, stolen Libyan arms have been trafficked and sold through illicit Lebanese, Turkish, and other sources into conflicts in Mali and Syria, and to bolster Hamas in the Gaza Strip. 43 Now, as Western powers such as the US and Canada prepare massive arms shipments to bolster Ukraine's defences, a similar fear is growing that those arms could be seized by pro-Russian paramilitary forces or Russian forces themselves, or sold privately through unconventional markets once the fighting subsides. 44
Dual use-technology complicating regulation
Another significant trend in the international arms trade is the advancement of dual-use technology and the globalized spread of co-production, both of which arose during the Cold War years. Dual-use technology “has both military and non-military applications, especially as the government application pertains to national security.” 45 Naturally, technological developments have advanced globally since the turn of the century, leading to more sophisticated and advanced arms. Twenty-first-century private-sector innovations in fields such as AI, nanotechnology, biotechnology, avionics, and advanced robotics have emerged as scientific breakthroughs promising significant impacts for healthcare, life sciences, and engineering. However, these advancements also pose security threats if modified for military use, especially those connected with possible weapons of mass destruction (WMD), or used in other forms of biological, chemical, or nuclear weaponry. 46
Pharmaceutical and biotech companies, for example, produce vaccines and medicines which are vital for human health. These same manufacturers also possess the means to produce toxic and biological weapons. 47 Specifically, fermenters used in manufacturing vaccines can be used to produce anthrax, while lab-based genetic modification processes can mutate existing pathogens or develop new ones. 48 This concern was heightened during the COVID-19 pandemic, during which countries scrambled to develop vaccines and scale up their biological and pharmaceutical capabilities. Hostile states such as North Korea, which likely has possessed a robust biological-weapons program since at least 2002, could conceivably use a pandemic as a means to acquire biotech equipment from the West or China through humanitarian pretenses, and then use the equipment to manufacture biological weapons. 49
Despite dual-use concerns with biotechnology in some countries, more regulated states have begun to scale back, and even destroy, some bioweapon stockpiles. The US began destroying its vast stockpiles after joining the Chemical Weapons Convention in 1997, along with 193 other states (excluding Israel, North Korea, Egypt, and South Sudan). 50 As of 2021, the US had only two sites tasked with the destruction of bioweapon stockpiles still in operation, with 78 percent of stores reportedly having been destroyed at the Pueblo Chemical Depot in Colorado, and 32 percent destroyed at the Blue Grass Chemical Agent-Destruction Pilot Plant in Kentucky. 51 Other UN conventions, such as the Biological and Toxin Weapons Convention, similarly prohibit member states from producing, transferring, and stockpiling biological weapons. 52 However, these conventions often lack inspection and enforcement mechanisms, leaving states to determine their own review mechanisms. For example, Canada voluntarily shares data on its laboratories, any research into biological defense conducted by the government and military, its capacity for vaccine production, and any legislation related to the BTWC. 53
Technologically advanced electronics and information systems are another dual-use concern. Satellites have allowed for the public use of global mapping, from location-tracking on smartphones to other GPS mapping services. Such satellite and tracking technologies can also be used as global surveillance systems or targeting systems, or for missile and drone guidance. 54 Further, these technologies are increasing the scope of cybercrime and the capabilities of non-state or hostile state actors. For example, technology seized by the Taliban during the American evacuation and fall of the Afghani government may give the Taliban access to US intelligence files and American encryption technology, which could allow the group or its partners to hack into public files belonging to the government or military, as well as private files of citizens and businesses.
Another example of a potential military application is the growing research and development of AI for use in the private sector. Though still in its infancy, there are dual-use concerns with AI as well. Regulatory complications regarding AI are discussed later in this work, but such concerns include biometric assassination systems 55 to target enemy combatants through armed drones and power naval “Ghost Fleets.” 56 Such technological developments complicate the international armament picture and remain the domain of non-state and surreptitious transactions. Nevertheless, some of the attendant technology and even finished products are exported and imported through company and government channels, and governments have keen interests in monitoring this traffic.
Monitoring technologies
As they become more advanced, dual-use technologies must be monitored and regulated in much the same way as conventional military arms and technology. One core regulatory concern is the need for “end-use” certification and enforcement. This is the practice of reporting the final destination and subsequent use or re-transfer of arms by the seller, typically to ensure that the arms are not used in violation of international law, in a conflict against an ally of the seller-state or its interests, or by hostile non-state groups. Though how these enforcements are applied can vary, the US monitors the end uses of sold arms by requiring the purchaser to refrain from trading or transferring the purchased equipment to a third party without written authorization by Washington, and to refrain, without authorization, from using the arms for anything other than expressed purposes. 57
Similarly, Canada governs its dual-use arms exports, as well as munitions and other military-related products and technologies, through the Export and Imports Permit Act (EIPA), through which an Export Control List (ECL) is created. Currently, the ECL contains nine groupings of “military goods and technology” of which the first pertains to dual-use technologies. 58 Citizens and businesses seeking to transfer goods found on the ECL must first register in the Controlled Goods Program (CGP) and pass a security assessment or background check. If an application for registration is approved, the individual or organization receives a certificate of registration which must be renewed after five years. However, transfers to the US, apart from firearms, nuclear technologies, and precursor chemicals, are usually exempt. 59
However, unlike the US, which conducts post-shipment checks through the “Blue Lantern” program, 60 Canada currently lacks well-defined systemic post-shipment verification processes. This lack of explicit end-use verification procedures, precipitated by an investigation into reports of Turkish armed drones engaged in conflict zones using Canadian targeting systems, prompted Ottawa to collect “best-practice” recommendations for post-shipment controls from other states that are party to the UN Arms Trade Treaty (ATT). Establishing such procedures is difficult, however, when a significant proportion of Canada's arms exports are component parts, such as targeting sensors, and not complete systems that can be easily identified and verified. Additionally, Canada's trade partners, such as the US, NATO, Saudi Arabia, and the UAE, would need to agree to on-site verification by Canadian officials, a move unlikely to be popular. 61 Despite these drawbacks, Ottawa took steps to prevent further end-use violations by canceling export licenses with Turkey for the components used in the Turkish TB2 combat drones, and stated that it would consider applications for arms transfers for NATO programs involving Turkey on a “case-by-case basis.” 62
Apart from complete weapons systems, some sensitive components, such as certain computer capabilities or advanced weaponry, have been subject to export controls for quite some time, as in past US restrictions on components for Saudi fighter jets (to allay Israeli concerns) or computer exports to China. Yet the challenges of keeping track of sensitive technologies have become even more urgent and complex. A recent analysis of seized Russian weapons systems and military components used in the invasion of Ukraine has revealed that at least twenty-seven Russian weapons systems rely on electronic component parts produced by Western powers and allies in Asia, including production companies based in Japan, South Korea, the UK, Germany, and Switzerland. More surprisingly, 318 of 450 identified electronic component parts are produced by fifty-seven companies based in the US, and include components such as signal processors and memory chips used in the guidance computer of the Iskander 9M727 “land-attack” cruise missile and transceivers used in the detonators of Kalashnikov KUB-BLA self-detonating “kamikaze” UAVs. 63
Despite export controls, many of these parts are thought to have been originally obtained under the guise of civilian purposes to circumvent arms controls and were later diverted for military use, or were acquired through “trans-shipment” methods, whereby a producer legitimately exports to a third party or state who then funnels the goods illicitly to the end user. In the case of Russia, many component parts that might otherwise be banned or restricted from the direct sale are sold to corporations operating in a third country, such as the King-Pai Technology or Sinno Electronics corporations based in Hong Kong, which then export to Russian corporations, many of which are currently sanctioned by the US government. 64 Highlighting the relevance of such trans-shipment methods, Italian customs authorities seized what were described as American drones, in transit to Qatar by way of Montreal, believed to be intended for final delivery to Russia, circumventing both American and Canadian military export controls. 65
However, increased Western sanctions—as well as increased Ukrainian capabilities to combat drones—may have limited the ability of the Russian military to effectively produce and deploy combat UAVs, as potentially signaled by the reported arms deal between Moscow and Tehran to train and provide the Russian military with armed drones. 66 Iran, having long faced trade restrictions and sanctions, is believed to produce its own UAV and armed drones without reliance on foreign-produced parts, 67 potentially allowing Russia to maintain its drone capabilities in light of sanctions and increased pressure on states to bolster end-use programs.
Another example of the risks of end-use violations and illicit arms transfers can be seen in American arms sales to Saudi Arabia and the UAE, ranging from conventional small arms and armoured MRAP vehicles to TOW anti-tank missile systems, which were later reportedly accessed by al-Qaeda and Salafi militias engaged with anti-government forces in Yemen. 68 These militias, loosely organized and accountable to no government, constitute an ongoing risk for uncontrolled violence, including alleged civilian attacks, child soldiering, and other war crimes. 69 Similarly, Canadian arms sold to Saudi Arabia have been found in use in Yemen, including LAVs and sniper rifles. 70 Canadian arms technologies and component parts have been used in other conflicts in violation of end-use agreements; for example, Canadian sensors were used in Turkish-made Bayraktar TB2 UCAVs (Unmanned Combat Aerial Vehicles), which were reportedly deployed in conflicts in Libya, Iraq, Syria, and Nagorno-Karabakh. 71 Regardless of the validity of the original transfer and the stated intentions of the recipient state, with any large-scale deal, there exists the potential risk of illicit resale and theft. One such possibility for end-use violation and illegal trade has developed due to the billions ($) in arms, especially explosives, missiles, and artillery systems that the US has transferred to Ukraine to assist in their defence against the Russian invasion. 72 Though no indicators of theft or resale have been uncovered thus far, Ukraine became a “trove for black-market arms trade” in 2014 when pro-Russian rebel forces seized Ukrainian military and police arms stores and pro-government militias were given arms with little oversight, leading to attempts to use foreign contacts to conduct sales. 73 Additionally, the loss of arms and weapons systems during combat against Russian forces could lead to their unintended use in bolstering Russian forces, as occurred during the American withdrawal from Afghanistan, 74 or they could be trafficked out of Ukraine and into other conflict zones, such as Syria and Yemen. In response to this threat, arms-control experts have begun planning for the possibility of illicit sales and trafficking. The European Union's “Border Assistance to Moldova and Ukraine” (EUBAM) program is meant, in part, to strengthen border security to prevent the trafficking of illegal arms, the use of GPS tracking of some transferred weapons systems, and Europol's announced commitment to work with Ukrainian officials to reduce arms trafficking in light of the heightened arms transfers to Ukraine. 75
Not only can failures of end-use agreements lead to civilian deaths, legal violations, and bolster the capabilities of hostile and unaccountable armed groups, but they can also create foreign policy blunders and harm the image of the exporting state. For example, when US arms were discovered in the hands of militia groups in Yemen, specifically the Salafi “Giants Brigade,” a pro-Iranian Lebanese television network broadcast the offloading of US armoured vehicles from UAE ships to a Yemeni port, increasing anti-American propaganda in the region. 76 Such use of American arms purchased legally by Saudi Arabia and the UAE would violate previously agreed-upon end-use conditions, and neither government received permission from the US to transfer the arms to any group fighting in Yemen. 77 In response, and in recognition of the ongoing humanitarian crises, the Biden administration announced an end of support to the Saudi military engagement in Yemen, 78 but announced no formal punishments for end-use violations, and after a brief hiatus continued with an arms deal to the Saudis, including the sale of F-35 fighter jets as discussed previously. 79
In response to reports of war crimes and abuses, other Western allies and Saudi arms trade partners began to scale back or ban further sales. Canadian slowdowns were noted previously, and Germany, in response to the humanitarian crisis caused by the Yemen war and the revelation that Western arms were reaching militia groups, passed a temporary ban on arms trades with Saudi Arabia and the UAE. 80 Italy cut off all military sales to these states—including blocking the sale of twelve thousand missiles—and the Biden administration temporarily froze sales to both states, 81 although the US ban was later lifted.
The diversion of Saudi- and UAE-purchased arms is not a novel controversy; both states have failed to file regular state export-import reports to the UN Register of Conventional Arms, 82 and states accused of end-use violations simply “fail to respond” to inquiries made by their trade partners; such was the case with Turkey's deployment of UCAVs equipped with Canadian sensors. 83
Implications of co-production
Co-production is also an increased security concern in the twenty-first century. Co-production is the agreement to transfer from one state to another the manufacturing technology or to jointly produce arms, including research and development programs, which in turn benefit both parties by lowering costs of research and manufacturing 84 and satisfying importing states’ concerns about costs, dependency, and employment needs. The most significant advantage of co-production is the reduction of development and production costs for the partnered states. By entering into co-production agreements, developing states gain not only the opportunity to bolster their military capabilities but also to counterbalance their more economically advanced neighbors. Co-production also allows developing nations to reduce their reliance on strong allies for military and security support, thus enhancing their perceived autonomy and capabilities domestically and in the international system. 85
Co-production can also be used as a foreign policy tool implemented by one state to positively change the arms regulation policies of the other. A 2017 co-production agreement between Sweden and Brazil to produce SAAB Gripen NG fighter jets, for example, was implemented with the secondary goal of causing a “general review of Brazilian arms export policies,” including developing “more transparent policies in line with the Arms Trade Treaty.” 86 Following this deal, Brazil formally ratified the ATT at the Fourth Conference of the State Parties in 2018. 87 Although Brazil's ratification cannot be explicitly linked to the Swedish SAAB co-production deal, it is quite likely that working with Swedish arms producers influenced Brazilian arms regulations and the desire to display a more controlled and transparent arms production and trade policy to the international community.
Another benefit of co-production is the ability for advanced industrialized states to work with strategic allies to help them build their own modern arms and armaments. This presumably helps the advanced state, which may be using co-production as a means to bolster the defences of a key ally as a form of off-shore balancing, 88 but it also helps the developing state, which may need to update its military capabilities and manufacturing technology, as well as create jobs, but cannot afford to do so adequately or responsibly through purchasing alone.
While co-production may be useful in lessening costs and as a tool of influence, it can also pose security risks. Similar to concerns regarding end-use, once the technology for a given weapon system is possessed by a state, their co-production partners cannot guarantee it will be used according to their own ethics, political interests or strategies, or international law. Further, co-production can become an issue when one state is included in the production and exchange of information but cannot maintain their payments, or when other states are excluded. One such example was the 2014 co-production of the South Korean-made KF-X fighter jet, a $7.9 billion partnership between Seoul and Jakarta in which Indonesia was to be held responsible for 20 percent of the budget. 89 In return, Indonesia was to receive forty-eight of the jets while agreeing not to sell the fighters or produce components using the technology themselves. Indonesia soon fell behind on their share of production costs, leading to a debt of $203 million by 2018 and a mutual agreement to re-negotiate Indonesia's role and contribution to the project. 90
The challenge of regulation, control, and accountability
To meet the shifting patterns, wider access, and advancing technologies of the international arms market, governments and international organizations must create and adapt regulations to monitor and record sales, implement “end-use” enforcement, and safeguard foreign policy and security interests. 91 As the world's leading arms producer and exporter, the US government is responsible for implementing many of these policies, for foreign military sales (FMS); private sales by American manufacturers to various parties; and direct commercial sales (DCS), which may come under legislative restrictions. 92 The American government values such sales programs in part to keep their manufacturers in business, lengthen production runs, thereby presumably reducing costs, and recoup government subsidies to the arms industry.
Foreign military sales are conducted between states in which one is seeking to purchase arms from the other's military supply. 93 In the US, a potential purchaser submits a Letter of Request specifying the desired military capability the purchaser is looking to procure and the approximate amount they are willing to spend. 94 After a period of review by the Department of State, and pending approval of Congress if the sale is greater than $14 million to a non-NATO state or $25 million to a NATO member, the sale is approved. 95
Direct commercial sales are those in which a purchaser, which can be a governmental entity or non-state actor, purchases arms directly from manufacturers. 96 In the US, this means that a foreign government can deal directly with an American arms manufacturer such as Boeing or Raytheon rather than the government. However, the exporter and specific sale must qualify under US arms export laws such as 22 US Code, subsection 2778, which grants the president power to control both imports and exports of “defense articles,” including deciding whether a given sale could contribute to “an arms race, aid in the development of weapons of mass destruction, support international terrorism, increase the possibility of outbreak or escalation of conflict,” or violate standing arms-control and nonproliferation agreements. 97
Canadian arms regulation exists within the EIPA and generally mirror the core conditions of the Arms Trade Treaty, specifically the considerations of whether the transfer (1) would contribute to peace or undermine it, and (2) could be used to commit or facilitate serious violations of international human rights law, including acts of terrorism, gender-based violence, or violence against women and children. 98 Additionally, arms transfers must be categorized by risk level. These levels vary between “low-risk” recipients—that is, like-minded states with similar export controls, such as the US—and those that are other than “low risk.” If a recipient is deemed to be other than “low risk,” the deal requires consultations and approval by members of Global Affairs Canada, the Canadian Department of National Defence, and other departments or agencies found necessary. If no risk is found, and the above criteria are satisfied, the deal must then be approved by the Ministry of Foreign Affairs. If the consultation concludes having found an issue or potential risk, it may still be approved by Foreign Affairs, even if a review committee recommends denial. 99 Vague interpretations of “risk,” as well as the Foreign Ministry's veto and override power, diminish uniform systemic controls of arms exports and allow Canadian arms to be transferred to states such as Turkey and Saudi Arabia, which have led to end-use violations and use in conflicts with reported human rights abuses.
The arms trade of the twenty-first century is a truly globalized industry, with cross-national co-production projects. Sovereign governments and militia groups alike may access conventional and complex arms, due to an increase in private manufacturing but also due to the ease of access through co-production partnerships, informal and black/gray markets, second-hand sales, and the spread of arms from failed or failing states. Thus, international organizations in addition to supplier nations have been tasked with implementing arms regulations to stem the flow of weaponry. Of these organizations, the UN has produced the most comprehensive arms regulation, the Arms Trade Treaty.
The ATT began—inefficiently—enforcing arms trades in 2014 and was designed to “prevent and eradicate the illicit trade in conventional arms and to prevent their diversion to the illicit market, or for unauthorized end use and end users, including in the commission of terrorist acts.” It aimed to “[establish] the highest possible common international standards for regulating or improving the regulation of the international trade in conventional arms.” 100 The ATT covers conventional arms in eight categories: battle tanks, armoured combat vehicles, large-calibre artillery systems, combat aircraft, attack helicopters, warships, missiles and missile launchers, and small arms/light weapons. 101
The states that have signed and ratified the treaty agree to regulate arms in these eight categories by establishing “national control systems” on export, including of parts and components (Article 4), providing a control list for public viewing to the UN secretariat (Article 5.4), and agreeing to prohibit any transfer of arms if it would violate UN arms embargoes (Article 6.1), international treaties, or agreements between the signatory states (Article 6.2), or if the state has knowledge that the arms would be used in commission of “genocide, crimes against humanity, grave breaches of the Geneva Conventions of 1949, attacks directed against civilian objects or civilians protected as such, or other war crimes” (Article 6.3). 102 Finally, if the transfer is not invalidated by the Article 6 prohibitions, ATT asserts that states must assess the potential that arms transfer might undermine peace and security, or be used to commit a violation of international humanitarian or human rights law, or terrorist acts (Article 7). If the state finds that any of these exceptions may be risked, they must not authorize the export. 103
The ATT has been approved and ratified by 110 states, including powerful exporters such as China and Canada, with another thirty-one states signed but having not yet ratified, including the three that have manufactured and deployed autonomous weapons systems—the US, Israel, and Turkey—as well as a major arms importer the UAE. 104 States that had neither signed nor fully ratified the ATT in the 2020s included Russia, Saudi Arabia, Egypt, India, and Iraq, several of which have been included in the top-ten arms importers since 2015. 105
For an international arms treaty to truly alter the flow of arms, it must impact the key importers and exporters. Those states, however, will not freely join a treaty that may hinder their sales (for the manufacturing, exporting, and re-exporting states), nor jeopardize their security, which they perceive to be at risk. Further, as the world's top arms exporter, the US is one of the most important states for such a treaty; yet, as a declining hegemon which remains the greatest power in terms of overall military capability, 106 it is unlikely to relinquish its sovereign authority and submit to international controls of its lucrative sales.
Washington has, however, joined the UN Register of Conventional Arms (UNROCA), which attempts to make the global arms trade transparent by collecting and documenting reported arms exports by type for each country reporting, as well as reported arms imports in order to cross-check for reporting accuracy. Washington is also notable for publishing, under statutory requirement, annual military expenditure, and arms transfer statistics for every country in the world. 107 The UN Registry suffers from its own deficiencies, however, and exporter/importer reports rarely match. Discrepancies may be due to the failure of a state to report in a given year, varying designations for weapons categories in different states, under-reporting by importers, damaged shipments, 108 or just plain evasion. For example, in 1998, the US reported sixty exported F-16 fighter jets to Turkey, which in turn reported receiving twenty-four. In 2016, China reported sending 636 missile and missile launcher systems to Pakistan—which does not report to the United Nations—making the exact number of Chinese missile systems possessed unknown. 109
Further, of the top fifteen arms-exporting states in 2020, five were not a party to the ATT, and combined they were responsible for 61.6 percent of the arms exported. This shows another weakness of the ATT and international arms treaties more generally; those states that are among the top suppliers, including the top two producers, decline to burden themselves with international law accountability regarding the lucrative sale of arms. Similar patterns exist with other international arms treaties, dealing with land mines, for example, and for the International Criminal Court. Without states such as the US and Russia, responsible for 58.8 percent of the arms sales of the top fifteen exporters, the ATT fails to impact the states most responsible for the flow of arms.
This is not to say that ATT members themselves are not responsible for arms trade and end-user agreement violations. As stated, Canadian arms sales to Saudi Arabia, as well as the use of Canadian imaging sensors in Turkish armed drones, have led the Canadian government to seek guidance from other members of the ATT to establish post-shipment controls. 110 France, one of the first states to ratify the ATT, also has been the source of significant Saudi arms imports, roughly €455 million in 2016, which were then reportedly used to arm Saudi-led militants in Yemen. Germany, another key ATT member, provided the Saudis with some €530 million worth of military helicopters, fighter jet parts and equipment, and radar systems 111 ; although Germany banned arms exports to Saudi Arabia in 2020. France has continued arming the Kingdom and has ignored calls from EU members to join a ban. Neither the Germans nor the French or other key suppliers have faced any form of fine or penalty for violating the ATT provisions prohibiting transfers if the state has knowledge the arms will be used “in the commission of genocide, crimes against humanity … [or] attacks directed against civilian objects or civilians protected as such.” 112
Conversely, China, by some estimates possibly the second-highest arms producer, ratified the ATT in 2020, and has positioned itself as the greatest power among those states party to the treaty. 113 This is significant as China has become a global power in the arms trade, but also because it exemplifies the shift of international power and influence. The US has continued to guard its sovereignty from international organizations, while China capitalizes on the US's absence by increasing its political clout. 114 Yet although it joined and ratified the ATT in 2020, China has yet to formally announce any changes in their arms production and sales, nor any changes in its laws regulating arms transfers.
The ATT faces issues, especially with enforcing violations. For example, Article 6.1 states that each state party is prohibited from making arms transfers in violation of UN arms embargoes 115 ; yet an arms embargo imposed on Libya since 2011 has been “totally ineffective” in stopping the inward flow of arms, 116 including from permanent UN Security Council member and ATT signatory France. 117
Circuitous transfer patterns also complicate regulation. Although French arms were not directly transferred to Libya, French “Javelin” anti-tank missiles and “Mirage” fighter jets have reached the Libyan National Army, led by Khalifa Haftar, by way of Egypt and the UAE—both of whom are key importers of French arms. 118 Egypt and the UAE are not a party to the ATT and therefore not subject to its prohibitions. Although clearly violating or skirting the ATT and the UN-imposed Libya embargo, France has not been subjected to any formal penalty, nor does the ATT specifically propose punishments for the violation but instead declares that disputes between states regarding the application of the ATT shall be settled through peaceful mediation, conciliation, negotiation, or adjudication (Section 19). 119
Rather, the most effective disciplinary action towards states with continuing controversial sales are legal suits brought by watchdog NGOs. 120 One such suit was brought by the Campaign Against Arms Trade (CAAT) against the British Secretary of State for International Trade. 121 The British Court of Appeals ruled in favor of CAAT, reinforcing the legal requirement of the British government to consider alleged human rights violations by Saudi Arabia in Yemen. 122 Although this ruling led to a temporary suspension of British arms sales to the Kingdom, it only served to reinforce the requirements to which states party to the ATT have agreed, rather than issue punitive discipline for unjust sales. Despite the temporary suspension in trade, the UK has continued its arms transfers to Saudi Arabia, doubling its trade of arms (in TIV) between 2020 and 2021. 123
In an attempt to bolster the ATT and its universal application, the EU Council on Common Foreign and Security Policy (CFSP) passed a decision in April 2021, which outlines a €1.37 billion plan to establish ATT national points of contact, create a roster of ATT experts to provide “advice and training on ATT implementation” at local and regional government levels, and to create a “database to match needs and resources” to link offers of and requests for assistance in treaty implementation. 124 This effort is meant not only to combat the costs of relying on international organizations (both non-governmental and those attached to the UN) to provide experts, including the costs of hosting such experts for training workshops, and the inefficiency that comes with relying on them rather than developing local and regional consultants, but also to establish long-term assistance for ATT implementation, albeit with little or no remedy for the unequal application of the law. 125
The future of advanced warfare and armament
Perhaps the greatest technological advancement to affect the international arms trade, and thus a trend for which arms regulation and control must prepare, is the development of AI for use in autonomous weapons. Once considered futuristic Hollywood fodder, AI, like many technological advancements, is now being adapted to control weapons systems and possibly to choose targets without human control.
An attack during the Libyan civil war in June 2021 was reported by the UN to have possibly featured a Turkish Kargu-2 drone programmed to engage targets without requiring “data connectivity between the operator and munition.” 126 This drone was deployed during a retreat by Tripoli-based government forces, though it is unclear whether the drone was able to engage an enemy through its own volition, or whether it was essentially a “loitering munition”—a drone deployed in a way that would lethally engage any moving human target—thus only useful during a retreat absent of soldiers from the deploying side. 127
A second instance of weaponized AI occurred in December 2020, when Iranian nuclear scientist Mohsen Fakhrizadeh was assassinated by an alleged Israeli weapon system that used an “intelligent satellite system” to target Fakhrizadeh in a moving vehicle. This application was purportedly precise enough to distinguish between Fakhrizadeh and his wife sitting in the seat beside him. 128 This use of AI seems to indicate a technology that can recognize a specific individual—in close proximity to others—through facial recognition, despite American assessments that show racial bias in facial recognition technology, including error rates as high as 34 percent for “darker-skinned” females. 129
Finally, the US Navy has received the next generation of unmanned naval vessels, dubbed the Sea Hawk, a predecessor to their first medium-sized Unmanned Surface Vessel (USV), the Sea Hunter. 130 These vessels are part of the Navy's “Ghost Fleet,” a project consisting of Large Unmanned Surface Vehicles (LUSVs) intended to be self-sustainable for ninety-day periods without the need for crew or maintenance, while possessing combat capabilities for electronic, anti-surface, and strike warfare. 131
It is evident that such advanced and still-experimental technologies are not likely to be traded or supplied abroad in the near-to-mid-term future by the advanced powers, since these technologies are highly proprietary. Yet it is unclear at what point that might change, especially within alliances such as NATO, and what modifications to export regulations might be leveraged.
These uses of weaponized AI mark a critical point in arms control and regulation. Although the current state of deployed autonomous arms and weapons systems lacks a level of sophisticated full autonomy, one in which a weapon system can freely differentiate between friendly soldiers and enemy combatants, as well as civilians, it does indicate that AI can be used to control weaponry and military combat vehicles. Further, such weaponry is not solely available to advanced industrialized nations; if an AI-controlled drone can be procured by the Tripoli-based military or militias in Libya, it might become available worldwide to other struggling governments as well as non-state groups. Additionally, as exemplified in the killing of Fakhrizadeh, AI-controlled weaponry may be used primarily as tools of assassination, a highly questionable legal tactic. Yet these systems and robotics in general are very tempting to states such as the US, which wishes to avoid “boots-on-the-ground” entanglements whenever possible.
These instances of weaponized AI are novel and remain irregular in the world of international arms transfers, but indicate a coming trend and consequent need for governmental and international regulations controlling the trade and use of autonomous weapon systems. The UN has been tasked with regulating “emerging technologies in the area of lethal autonomous weapons systems” and mandated to produce guiding principles for implementing new international laws in line with the UN Convention on Certain Conventional Weapons, despite pushback from Russia and fruitless deliberations which have been ongoing since 2014. 132
One such concern, highlighted often in science-fiction literature, regards human control, that is, the need to ensure that humans remain in charge of the application and decision-making of autonomous weapons so as to conform to international law and ethics. 133 In order to meet the goal of regulating control of autonomous weapons, governments would have to implement measures covering: (1) the interaction between the human controller and the machine, including measures which allow for human intervention; (2) the environment in which an autonomous weapon is used, such as restricted use on civilian property and in areas with civilian populations and non-combatants; (3) the type of target the autonomous weapon is used for, including the ability for immediate deactivation by a human controller. 134
AI, even with the capability to “think” and act independently of human control, and even if regulations and pertinent international laws are programmed in, may not be sufficiently discriminating in all cases. Some nations have begun considering the legal status of AI, and their decisions may shape the future of AI and the law. For example, in 2017, Saudi Arabia granted citizenship status to an AI-powered robot named “Sofia”; the same year, Japan granted citizenship to a “chat bot” named “Shibuya Mirai” under a special regulation. 135
Not only are these instances significant and quite bizarre in potentially laying the groundwork for a more recognized system of applying law to AI, but each case saw a state subvert its own citizenship laws in order to grant status to machines. 136 Though it is possible that these cases were more about public displays of technological achievement than making both AI and humans subject to the same laws, they introduce an inequitable application of the law, which may become more problematic as AI is transitioned for military use. 137
On 7 March 2022, delegations from the US, Canada, Australia, Japan, the Republic of Korea, and the UK met as the Group of Governmental Experts (GGE) on Emerging Technologies in the Area of Lethal Autonomous Weapons Systems. The purpose of this meeting was to submit proposals for “Principles and Good Practices on Emerging Technologies in the Area of Lethal Autonomous Weapons Systems,” meant to guide future legally binding policies under international humanitarian law. 138 Although this indicates a growing priority among advanced nations to monitor and regulate militarized AI as it becomes more commonly used, it displays a general hesitancy to formalize any international regulations. Indeed, the meeting bore the fruit of “good practices” and voluntary principles, rather than limits or regulations, leaving the possibility for the top arms manufacturers to continue to adapt AI for military use without restriction. 139
Conclusion
The state of the international arms trade is rapidly expanding and evolving, and the international community, through the UN and other organizations, is continually seeking regulations to report or stem the flow of arms and prevent the human rights abuses which may be linked to arms transfers. 140 Attempts at regulation are reflexive and must adapt to advances in technology and changes in the environment of arms production and trade. Rising powers and regional leaders seek to expand their military capabilities, maintain job creation, and increase financial gain, and are doing so by investing in their own arms-manufacturing capabilities, as is evident with Canada's continued dealings with Saudi Arabia and Israel, 141 the rise of China in the arms trade, 142 and ventures into high-tech arms manufacturing by regional powers such as Israel, Brazil, and South Korea. 143
While ATT signatories such as China and Canada, though patterning regulations to fit the treaty, continue to avoid punishment for arms transfers to states that are linked to human rights abuses and attacks on civilians, other states are emphasizing the need for more efficient and enforceable arms regulations. Germany has led such EU-wide efforts, 144 and the CFSP is working on implementing projects to make ATT enforcement more efficient throughout Europe, which might lead to stricter enforcement efforts regarding members such as France. 145 Canada, attempting to maintain its posture as an international peacekeeper and law-abiding arms dealer, has sought the aid of ATT member states—primarily Germany—to identify possible post-shipment end-use verification measures to prevent future violations, such as those reportedly committed by Turkey. 146 Successful end-use monitoring remains one of the most critical and necessary goals of contemporary arms regulation, especially as the number of arms producers continues to grow. Hostile states or sanctioned such as Iran can independently produce and deal combat drones regardless of international arms-control efforts, and transgressors of standing international treaties and arms regulation continue to evade punitive consequences for arms deals which led to end-use violations.
AI has not yet been fully developed or adapted to reach its maximum military utility, and international arms regulatory bodies are challenged to create enforceable legislation for these advancements. Agreements such as the Chemical Weapons Convention 147 and Biological and Toxin Weapons Convention 148 have implemented the controlled destruction of dangerous and difficult to control biological agents such as mustard gas, and have seen countries such as the US significantly diminish their stockpiles while others such as Canada self-impose non-mandatory compliance mechanisms to maintain transparency. 149 This tends to happen, though, when several countries gain the technology and threaten each other in deterrence relations. Although difficult to enforce worldwide, especially in more hostile states such as North Korea, the controlled destruction of such agents is promising.
Other destabilizing, widely possessed, and indiscriminately destructive technologies such as land mines and cluster munitions have generated international control agreements, such as the Ottawa Treaty banning the use and transfer and requiring the destruction of anti-personnel mines. Notably the US, Russia, and China, as major military powers, have not formally adhered to the agreement, but claim to abide by some provisions and controls such as requiring self-destructive mechanisms, destroying certain stockpiles (for the US not including mines for the Korean demilitarized zone and for Russia not including their use in Ukraine), and limiting exports. 150 Nevertheless, some of their reporting has been vague about such adherence. Thus, much work remains to strengthen the international control regimes.
The conflicts of tomorrow may be those in which humans, both military and civilian, are targeted as the result of AI programming and not the direct use of force by another human. These changes must be met by the international community and the US as the world's leading arms dealer to stem and control the flow. Regulations such as the ATT seek to create a globally agreed-upon and transparent set of conditions and prohibitions governing arms trade, but without major exporters—the US, France, China, and Russia—as a party to and adhering to the provisions, and moreover, with unequal enforcement of violations committed by ratified members, stemming the global licit and illicit flow of arms, especially to factionalized conflicts such as Yemen and Libya, will be highly problematic.
To effectively monitor the globalized sale of arms and military components, arms regulations need to be implemented and enforced by international bodies with the power to apply regulation, restriction, and penalties equally among all member states. Yet the efficacy of these measures wanes when enforcement of violations is lacking and top arms producers are not a party to the regulations. International bodies like the UN have not yet taken punitive measures, and court rulings on behalf of NGOs have only led to short-term transfer suspensions. Lucrative arms transfers to allies—especially those facing dire security threats—and other clients will remain tempting, and of course, involve private arms merchants supplying conflict zones in many regions. End-use agreements and transfer report transparency will remain elusive but imperative, if for no other reason than the desires of major powers to maintain control over who becomes armed, to what extent, and for what purposes, especially in response to hostile military actions.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
Author Biographies
Layton J. Mandle is a doctoral candidate at Wayne State University. His research interests include world politics, theories of war, defence policy, and arms trade. His doctoral dissertation research is a quantitative analysis of world leadership capabilities. He currently works as a part-time lecturer at Eastern Michigan University.
Frederic S. Pearson is a professor of political science and former director, the Center for Peace and Conflict Studies at Wayne State University. He is a recognized authority in the fields of international military intervention, arms transfer effects on wars, and civil and international conflict analysis. He has been recognized at Wayne State University as Gershenson Distinguished Research Professor and as Graduate Mentor of the Year.
