Introduction
‘Third countries’ are frequently exploited by those involved in networks to transfer proliferation-sensitive technologies. This allows procurement agents to obscure the end user or vendor located in the proliferating state, and to avoid arousing suspicions of industry, export licensing officials, and intelligence services. While ‘third countries’ have frequently featured in illicit transactions, the academic literature exploring the roles played by entities in these jurisdictions is limited. This article uses an in-depth case study of Malaysia, a country having frequently played a role in these networks, to provide insights into ‘third country’ selection by procurement agents or arms traffickers. Rather than seeking to single out Malaysia, the article uses the country – one of a number that have prominently played a ‘third country’ role in proliferation networks – to explore how these countries are used by these networks, and the behaviour of individuals and entities of which they are composed.
Building on the literature on economic sanctions busting, the article presents a loose typology of the ways these networks use third countries, and then illustrates its applicability using three detailed case studies involving Malaysia-based entities: A. Q. Khan’s nuclear black market network, and Iran and North Korea’s efforts to procure and market WMD-related and military goods. In doing so, the article considers what factors are involved in how proliferators select ‘third country’ hubs. The conventional wisdom suggests that Malaysia and other jurisdictions such as the UAE have featured prominently in these networks because of weak export controls, regulation, and oversight. The article compares proliferation networks’ exploitation of third country hubs to Bryan Early’s discussion of their use in trade-based sanctions busting.Footnote 1 It is argued that while distinct phenomena – notably trade-based sanctions busting focuses on volume of trade, while small, high-value transactions can make a big difference in proliferation networks – third country hubs in proliferation networks share some characteristics as hubs of trade-based sanctions busting. The cases explored in the article also suggest that factors beyond the lax regulatory environment, such as the level of development, and personal connections, are often as important in driving the decisions of proliferation networks. The article concludes by considering implications for nonproliferation policy.
Nonproliferation, strategic trade controls, and ‘third countries’
Throughout the latter half of the twentieth century, efforts have increasingly been made to prevent the proliferation of WMD – nuclear, chemical, and biological weapons – and their means of delivery. As these efforts have seen a reduced number of governments willing to provide WMD-related technologies to other states, proliferators have increasingly turned to the international marketplace to obtain technology for their weapons programmes. As the main supplier states have put in place export control systems in order to minimise the risk of diversion of exports to WMD programmes, those seeking to procure WMD-related technologies have increasingly used illicit procurement networks and techniques to obtain controlled technologies.
Recent nuclear and missile proliferation cases – Iran and North Korea – have seen multilateral sanctions imposed against them by the UN Security Council, as well as unilateral sanctions imposed by various states. These complex sanctions regimes have included technology and arms embargos, travel bans, and asset freezes imposed on those directly associated with the weapons programmes. The most recent North Korea resolutions have also imposed sanctions on sectors of the North Korean economy.
The rationale behind these technology-based sanctions on nuclear and missile programmes – so-called ‘supply-side’ measures – has been to prevent Iran and North Korea from obtaining requisite technologies, to slow the programmes’ development and to raise their costs. Under UN Security Council Resolution (UNSCR) 1540 (2004), countries have been legally mandated to put in place export controls and other systems, in theory allowing them to implement sanctions, although UNSCR 1540 implementation has been very patchy in practice.
Economic sanctions more broadly – including the recent sectoral elements of UN North Korea sanctions – have been intended to affect the ‘demand-side’, pressuring Iran and North Korea to halt their pursuit of nuclear technologies. The UN arms embargoes have also sought to affect these countries’ will and ability to pursue nuclear weapons; arms sales have functioned to bolster ideological and political connections with allies, and in the case of North Korea, raise hard currency for its nuclear programme.
To circumvent export controls and breach supply-side controls, procurement agents working for Iran and North Korea have employed techniques to deceive intelligence services, export controllers, customs officials, and industry compliance officers that are seeking to prevent exports to WMD programmes. They include, but are not limited to, use of front companies, falsifying documentation, and concealing or mislabeling shipments. To breach the UN arms embargo, North Korea has also used front companies to market its arms. While many of these techniques remain unchanged since the 1970s, observers have suggested an increasing sophistication over time as procurement networks adapt to expanding measures to prevent proliferation.Footnote 2
One of the techniques most frequently used by proliferators is placing procurement agents or brokers, or routing shipments, through third countries. The term ‘third countries’ or ‘third party state’ is used in the literature on economic sanctions to describe players that are not the sanctions ‘sender’ or the sanctioned or ‘target’ state.Footnote 3 As used in this article, the term has a similar, but distinct meaning. Because some of the cases involve UN sanctions (universal and legally binding on all countries, which in theory means there is no ‘third party’) the term is used to describe countries utilised in transactions that are not the origin or destination of the goods being transferred. Other terms have been used in this manner, including hubs for ‘transshipment’ or, in cases of procurement, ‘diversion’ (where goods are diverted from their intended destination) and ‘turntables’ (where goods are imported and immediately turned-around and re-exported).Footnote 4
While the use of third countries for deception purposes has frequently been seen in WMD and military goods proliferation networks, there has been little effort to conceptualise the role played by these jurisdictions in the scholarly or policy literatures. Explanations have tended to focus on third countries being selected on the basis of weak export controls and enforcement. More nuanced conceptual thinking on this topic could have great value for our understanding of proliferation networks, and for developing proactive policies to counter them.
The proliferation networks and sanctions busting literature
The existing scholarship largely falls into two different areas: Literature on the role of third country hubs in networks trafficking military and WMD technology has been fairly limited. There remains scope to further our understanding of the role of third countries by building on a second area of literature that is more developed: that exploring the role of third countries in economic sanctions busting.
References to ‘third countries’ in the proliferation network literature – that focused on the transfer of technology, often in breach of export controls or sanctions – are generally cursory and made in passing.Footnote 5 The literature on proliferation networks has largely focused on the state-level, with states forming the nodes: both suppliers and recipients.Footnote 6 More recently, some work has started to address what could be described as the ‘transactional level’, focusing on the role of, and interactions between, organisations, companies, and individuals.Footnote 7
This work has sought to consider why elements of the private sector – individuals and companies – become involved in WMD-related illicit trade, providing a framework through which key motivations of profit, political, and ideological interests can be considered.Footnote 8 However, while it is the private sector that largely provides the technology to proliferation networks, often the middlemen who broker these transactions, alongside the needs of these procurement networks, are ultimately driven by the state programmes they supply. These ‘witting’ aspects of proliferation networks, especially the aspects based in third countries, have yet been fully explored in the proliferation network literature.Footnote 9
Justin V. Hastings provides the most developed conceptual treatment of third countries, taking a geographical approach to the Khan network, where nodes are ‘people or organizations anchored in a specific piece of territory’.Footnote 10 He argues that networks without state prerogatives or resources – such as state-owned or military transportation, or embassies – must ‘set up support structures that depend on advantageous economic, political, and social characteristics of their host countries’.Footnote 11 Considering Khan’s efforts to supply Libya, Hastings focuses on Dubai as a key third country hub, as well as briefly considering the manufacturing operations in Malaysia, providing insights into selection of these hubs. Dubai’s ‘political and economic environment’, lax regulations on setting up companies, place in global transportation infrastructure, high proportion of international residents, and lack of government oversight; and Malaysia’s technological sophistication, lax regulation, and Khan’s associate’s existing social network are all cited by Hastings as factors.Footnote 12
Hastings’s valuable work is not without limitation, essentially drawing its insights from three interlinked networks – all involving A. Q. Khan and some other common actors. The Khan network, arguably the most well known and damaging illicit non-state supplier network in history, is clearly an anomaly, and possibly an anachronism. As Stephanie Lieggi has noted, while Hastings portrays lack of state resources as a limitation, in recent years networks have moved from reliance upon state resources to take ‘greater advantage of the globalized trade system’, especially through benefiting from ‘the virtual anonymity that can come with the use of major transshipment hubs and manufacturing locations’.Footnote 13
The scholarship on economic sanctions busting, generally more developed than that on technology embargoes, has considered the role of third countries in undermining sanctions, but has not been consulted in relation to proliferation networks. The term ‘black knights’ was coined to describe ‘powerful or wealthy countries’ that provide support to undermine the effects of economic sanctions.Footnote 14 More recently, Early has drawn distinction between politically driven ‘aid-based’ and opportunistic profit-driven ‘trade-based’ sanctions busting, his work providing arguably the most nuanced treatment of sanctions busting to date. Early’s concept of trade-based sanctions busting involves the development of ‘alternative trading relationships’ driven by profit-seeking private sector actors.Footnote 15 This is more relevant to considering proliferation networks than ‘aid-based’ sanction busting, which – driven and managed by governments – has more similarity with the state-to-state proliferation-related transfers.Footnote 16 Early’s focus on economic sanctions necessarily means that his work concerns third parties spoiling sanctions through making up declining trade with large volumes of business transactions. Focus on volume, or including proliferation-related alongside economic sanctions busting, is seen frequently in the literature.Footnote 17 While there are some similarities in the trade-based sanctions busting he has conceptualised – notably that goods tend to be sourced from the private sector and profit constituting the driver in some cases – the behaviour of proliferation networks, in which small numbers of specialised transfers can make a big difference, is a different phenomenon worthy of similar nuanced treatment.
In sum, the nascent conceptual literature on proliferation networks has only considered third countries in passing, while that on sanctions busting largely considers a similar but separate phenomenon. Discussion has failed to systematically consider the roles that these jurisdictions can play, and what factors results in their selection by proliferation networks. This article uses a number of in-depth case studies in a single country – Malaysia – to consider questions specific to the role of third countries.Footnote 18 The Malaysian case presents a valuable opportunity for inquiry. The state has had no interest in developing WMD, and has a fairly small-scale but expanding defence industrial base. However, the country has seen significant proliferation-related activity over the past two decades through the Khan network, and Iranian and North Korean illicit trade, with many details about these activities in the public domain. Cases considered below include transfer of nuclear and missile technologies, as well as military goods and US origin technologies covered by the Iranian embargo. Emphasis has been placed on extracting data relating to individual decision-making – for example, through emails quoted in court documents – when available. In some of the Iranian cases involving goods covered by the US embargo, the goods themselves are fairly benign.Footnote 19 While the application of export controls can be politically and ethically contentious, this article avoids discussion of these questions, focusing on the behaviour and decisions of the networks. The significant data surrounding the cases allows for extraction of data points directly relating to decision-making by individuals involved in these networks, yielding fresh insights.
Third countries in proliferation networks: Roles and rationales
Given the paucity of existing conceptual scholarship on proliferation networks, this section presents a loose typology of the roles that third countries can play. Entities based in third countries can play three main roles in proliferation networks (Figure 1). This typology encompasses the vast majority of scenarios, and with the latter two roles – transshipment and brokering – being much more commonly seen than the manufacturing role.Footnote 20 Each of these roles will be illustrated using a Malaysia-related case study in the subsequent three sections.
Figure 1 Typology of third country roles.
In presenting this typology, three caveats are stated: First, the typology largely refers to those that knowingly decide to utilise third countries in proliferation networks, and are ‘witting’ actors – those that understand that the goods are destined for a WMD programme.Footnote 21 Second, the typology above does not include ‘transit’ – this is where goods might pass through the country’s ports, without being offloaded from the ship. This is because, rather than being indicative of the country being utilised in a proliferation network, cases where goods have transited a port tell us more about global logistical routes and transportation networks, and cases where shipments are seized while in transit tell us more about interdiction or enforcement action.Footnote Footnote 23 Third, the typology focuses on activities involving the movement (or planned movement) of goods, rather than also considering enabling functions such as financing or transportation.
The three third country roles set out have different ‘footprints’ – in spatial, social, legal, and bureaucratic terms. The footprint is seen in a physical sense – manufacturing requiring a factory or similar facility; transshipment requiring a basic office set up, and possibly some kind of warehouse for larger items; and marketing or brokering requiring little more than a phone or Internet connection.Footnote 24 In this context, footprint is also used to encompass often observable legal, bureaucratic and social, impact of these activities. A large factory or warehouse involved in a proliferation network could provide significant physical evidence of a network’s activity. However, it also requires satisfying (or avoiding) significant legal and bureaucratic processes – everything from insuring imports and exports clear customs, employing and managing skilled labour from the local community or sourced from other countries, acquiring and maintaining manufacturing equipment, to paying corporate taxes.
A significant legal dimension relates to whether the actor decides to legally register a company in the third country to undertake its activities, or whether the illusion of a company is used. Creating this illusion could involve using a letterhead or a website. Both pathways create challenges and opportunities. The barriers to establishing a legal entity vary between jurisdictions. The legal registration may satisfy those conducting in-depth due diligence, but could also involve more information on the organisation of benefit to investigators being placed into the public domain.
On the other end of the spectrum, the footprint of an individual operating as an arms broker could be minimal: operating alone, conducting meetings in hotel bars, carrying just a laptop and a phone. Individuals could, and frequently do, broker deals involving goods being shipped between two separate jurisdictions, with goods never passing or being transshipped through the country in which they are located. Footprint, as used here, does not necessarily imply the ability of the third country government or others to detect the proliferation-related activity. This would depend on the network’s ability to hide its activities, and operational security.Footnote 25
The use of third countries could in theory be avoided in all three of these transactions. Goods could be procured directly from a supplier by the country hosting the WMD or military programme, without need for transshipment or a broker located in a third country;Footnote 26 the function fulfilled by a third country factory could potentially be replicated in the destination, or more likely in the advanced supplier economy. However, the main rationale for the use of individuals or entities based in third countries is to deceive – to hide the ultimate end-user of goods, connections to a sanctioned country or programme, or in the case of the factory to prevent the need to procure finished goods with clear WMD application.Footnote 27 This avoids raising concern among industry and governments that would be triggered by direct approaches from, or attempts to license exports to, Iranian or North Korea-based companies. In doing so, the use of a third country adds a layer of deception by means of exploiting what is a normal characteristic of most international supply chains.
When it comes to the selection of specific hubs, the literature on criminal networks has coined the term ‘jurisdictional arbitrage’ to refer to transnational criminal groups efforts to ‘exploit the differences in national laws and regulations’.Footnote 28 Countries that offer an environment with limited business regulation, limited enforcement activity, and related oversight – with regard to export controls and in other respects – would clearly be advantageous to those seeking to avoid detection and disruption of their activities. Lack of political commitment on the behalf of the host government to the implementation of sanctions or export controls and a significant diaspora business community of the sanctioned country may also be advantageous.
In the geography of illicit networks, distinction has been drawn between terrorist ‘havens’ offering lawlessness, and criminal ‘hubs’ providing some ‘baseline level of infrastructure and services’.Footnote 29 The manufacturing and transshipment or re-export operations in the typology would clearly benefit most from those features of ‘hubs’: access to commercial transportation routes, secure warehousing, and in the case of a factory uninterrupted supplies of water and electricity, and access to skilled labour. The solo arms broker example would not necessarily require these features as urgently, although true ‘lawlessness’ would not be beneficial. The necessity of these ‘hub’ attributes is both to allow the basic functions of a factory, or warehousing, but also because requests for quotations for highly advanced dual-use technologies originating in countries without a sufficiently developed industrial sector are likely to raise alarm.
The next three sections present detailed case studies that illustrate these three types of third country roles. The cases are designed both to showcase the applicability of the typology, and to consider why the proliferators in those cases chose particular third country hubs. The cases are followed by a section that seeks to draw broader conclusions about the choice of Malaysia and the behaviour of proliferation networks, before considering recommendations for policy.
Khan, SCOPE, and Malaysia
In the early 2000s, the scale of the proliferation activities of Pakistani nuclear weapons scientist A. Q. Khan started to become clear, although the full story regarding the Pakistani state’s knowledge of his activities remains unresolved. Drawing on many of the contacts Khan had established in procurement for Pakistan’s centrifuge enrichment programme, Khan went on to coordinate the supply of centrifuge and other technologies to Iran, North Korea, and Libya. The ‘Libya deal’, struck in 1997, involved the supply of a full gas centrifuge plant.Footnote 30 Fulfilling such a large order would require a more extensive manufacturing capability than previous deals that mostly relied on surplus goods from Pakistan’s programme.
Malaysia was selected by Khan’s operation to host a significant manufacturing effort. Scomi Precision Engineering Sdn BhdFootnote 31 (SCOPE) was used to manufacture centrifuge components. Four shipments were made of these components to Libya. The final of these shipments, which included 25,000 centrifuge parts – labelled ‘agricultural machinery’ in SCOPE marked crates – was shipped from Malaysia on a local ship in August 2003.Footnote 32 The shipment was warehoused in Dubai for 48 hours before being transferred on to the BBC China, which was interdicted in Italy on route to Tripoli.Footnote 33 This first case considers Khan’s manufacturing operation in Malaysia and the rationales for the establishing operations in this jurisdiction.
The Libya deal was on a scale that hadn’t been dealt with by the network before. It would eventually involve the transfer of 10,000 P-2 centrifuges – each including around one hundred parts and components – meaning the production or procurement of around one million parts.Footnote 34 The clear scrutiny any effort to locate new manufacturing operations in Pakistan or then-sanctioned Libya would garner led Khan and B. S. A. Tahir – a key figure in Khan’s network – to consider other options.Footnote 35 The most complex centrifuge parts were manufactured in the a factory in Switzerland owned by long-time collaborators, the Tinner family, while production of less complicated parts was outsourced to other ‘third countries’ – Malaysia, alongside South Africa, Turkey, and others.
Accounts of the decision to establish operations in Malaysia suggest that several alternatives were considered. Originally, Dubai had been considered as an option – Urs Tinner had sought to establish a factory there but was unable to find a sufficient local skilled work force, and worried that work permit applications may stimulate government interest.Footnote 36 Second, Turkey was considered, but again the lack of skilled labour proved to be a problem.Footnote 37 A third location, South Africa, was also considered. South Africa had hosted a nuclear weapons programme that was dismantled around 1990, meaning labour shortages were less likely to be an issue. However, South Africa’s history as a proliferator meant that governments may be wary about shipments of goods such as maraging steel to the country.Footnote 38
Malaysia was first flagged as a further option by B. S. A. Tahir in mid-2001, with the option explored further in the autumn.Footnote 39 Tahir, a Sri Lankan businessman, ran the Dubai hub, and rapidly became Khan’s right-hand man.Footnote 40 Tahir had several personal and professional links with Malaysia. In June 1998 he married a Malaysian woman – Nazimeh Syed Majid, daughter of a prominent Malaysian diplomat. Khan, alongside other network members, attended the wedding just weeks after Pakistan’s first nuclear test.Footnote 41 Tahir’s marriage meant that he was eligible for permanent residency in Malaysia, although he generally spent most of his time in Dubai, as the Police report noted, only returning to Malaysia ‘once in a while, to visit his wife’s family or look for business opportunities’.Footnote 42 However, by 2000, Tahir owned an expensive building in Kuala Lumpur, and other Malaysian businesses.Footnote 43
Tahir ‘mixed with Malaysia’s elite’ and grew close to Kamaluddin Abdullah, the son of Abdullah Ahmad Badawi.Footnote 44 Badawi was a long-time Malaysian MP, and would become deputy prime minister in 1999, and prime minister in 2003. Kamaluddin appointed Tahir a Director of his privately held investment company Kaspadu, which controlled Scomi Group. Tahir’s wife was also an investor in Kaspadu, and served on its board after Tahir.Footnote 45 While there were clear high-level connections to the Malaysian establishment, there is no evidence that Kamaluddin was aware of the nuclear dimensions of Tahir’s activities.Footnote 46
The decision to move operations to Malaysia was allegedly precipitated by a break-in at one of the network’s Dubai warehouses.Footnote 47 Scomi Group signed a two-year $3.43 million contract in December 2001, and an existing company was acquired to handle the contract and renamed SCOPE.Footnote 48 Urs Tinner – by this point working as a CIA informant – moved to Malaysia to work as a consultant and help establish operations.Footnote 49
An existing factory at Shah Alam outside of Kuala Lumpur – now owned by SCOPE – was refitted and would host thirty workers. Its capability was upgraded from producing car parts and industrial tubing to centrifuge components.Footnote 50 The upgrade, and the factory’s new operations, required the import of equipment and materials. Aerospace grade aluminum was provided to SCOPE by Bikar Metal Asia in Singapore, sourced from entities in four other countries – Germany, Russia, Slovakia, and Italy.Footnote 51 Machine tools were procured from European and Japanese suppliers.Footnote 52 Those working in the factory were allegedly unaware of the planned nuclear end-use for the products, as Urs Tinner took special efforts to safeguard all relevant documentation.Footnote 53
The Khan case provides a starting point to consider the reasons for the exploitation of Malaysia by proliferators. Catherine Collins and Douglas Frantz note: ‘Khan was intrigued because Malaysia offered a good technical base, lax export controls, and a location far from the spies and customs authorities of Europe and the United States.’Footnote 54 In 2004, Malaysia did not have a comprehensive export control system in place, and SCOPE’s activities did not break Malaysian law. Malaysia was also a location less likely to cause concern among those supplying technology for the SCOPE venture.
Hastings supplements the notion of limited regulation and a sufficient technical base with ‘a social network to ensure the operation’s success’.Footnote 55 He states: ‘The choice of Malaysia … illustrates how Khan’s network was constrained by the need for social ties and the political and economic characteristics of the countries in which it operated.’Footnote 56 He notes the centrality of Tahir in decision-making, his residency status, shares in Scomi, and political connections.Footnote 57 After the BBC China interdiction, Tahir left Dubai for Kuala Lumpur, allegedly expecting that ‘his political connections would protect him’, while Tinner packed up and fled Malaysia seeking to ensure little evidence was left behind.Footnote 58 These factors suggest the importance of personal connections and circumstances in third country selection.
In 2004, as a product of the Khan Network and broader concerns regarding WMD terrorism, the Security Council passed UNSCR 1540, requiring states put in place a series of measures to prevent WMD proliferation – including export controls, border controls, and other measures. In its first 1540 report, the Malaysian government noted it lacked a comprehensive export control law.Footnote 59 New comprehensive legislation would not be put in place until 2010.
The SCOPE case suggests the choice of Malaysia was shaped by needing a location with lax regulation, a lack of oversight, and a supply of skilled labour. This saw Malaysia considered over Dubai and Turkey (insufficient workforce) and South Africa (perceived oversight). However, other countries could potentially have fulfilled these criteria. Existing personal connections of B. S. A. Tahir and his belief that the country could provide political ‘cover’ played an important role in the selection of the country.
Iranian procurement activity in Malaysia
In the late 2000s, Malaysia was an important transshipment hub for Iranian illicit procurement activity, with Iranian agents operating in the country to procure military and missile related goods. Leaked US State Department cables, US court documents, and other sources allow us to piece together a picture – albeit a fragmented one – of these activities. Malaysia has featured prominently in studies of Iranian illicit procurement. A US Government Accountability Office study noted that Malaysia or Singapore was involved in 20 per cent of cases of Iranian illicit procurement of US origin military and dual-use goods.Footnote 60 This was only surpassed by the UAE, which was involved in around 50 per cent of the cases.Footnote 61 Another study presents slightly less evidence of Iranian use of Malaysia in nuclear-related illicit procurement.Footnote 62 However, Iranian activity in Malaysia in the late 2000s saw transshipment of electronics of use in Improvised Explosive Devices (IEDs), and aerospace and missile technologies.Footnote 63
Following the coalition invasion of Iraq in 2003, Improvised Explosive Devices (IEDs) constructed using Iranian supplied components, were used extensively and claimed many casualties in the resulting insurgency. From at least 2006 until 2008, entities based in Malaysia were involved in a global procurement network sourcing key electronic components used in IEDs.Footnote 64 Malaysia-based and Iranian-operated, Vast Solution Sdn Bhd was involved in at least 12 procurements from US companies. These procurement efforts were linked to those of Dubai-based Mayrow General Trading Company, which had been operating since at least 2004. Procurement through the ‘Malaysia conspiracy’ increased after the efforts of Mayrow in Dubai seemed to tail off around 2006, with the Malaysian network gradually replacing the role of the Dubai network.Footnote 65 Vast Solution utilised direct Iran Air flights to transfer the goods to Tehran.Footnote 66
In the same period, a wider network of Iranian front companies and procurement agents operated in Malaysia. In 2006, the UN Security Council put in place the first technology-based sanctions on Iran’s missile programme.Footnote 67 At least ten entities and eight middlemen were operating on behalf of Iran in Malaysia during the 2008 to 2010 period.Footnote 68 These entities also operated in close connection to at least four in neighbouring Singapore.Footnote 69 The US government noted this expanding Iranian activity in a 2008 paper on ‘proliferation trends’:
Over the past several years, companies in Malaysia repeatedly have attempted to procure a variety of aerospace-qualified electronics from the U.S. and other MTCR Partner countries on behalf of military- and missile-related end-users in Iran. It also appears such companies … are expanding their procurement operations, regularly using multiple cover names and fraudulent end-user documentation, and routing their transactions through additional intermediaries to conceal the ultimate destination of an export.Footnote 70
These Iranian procurement networks operated in two interlinked clusters.Footnote 71 The first, was a cluster linked to the Iran-based Farazeh Equipment Distributor Company (FEDCO), an Iran-based supplier of SHIG (Iran’s liquid fueled missile programme) and the Iranian Unmanned Aerial Vehicle (UAV) programme.Footnote 72 FEDCO has been described as a ‘parent’ company of Malaysia-based front companies Evertop Services Sdn Bhd and Elite Advanced Solutions Sdn Bhd.Footnote 73 FEDCO also employed Malaysia-based middlemen and brokers, including one that was seeking a data acquisition system of use in UAVs or satellites from a Belgian company.Footnote 74
Evertop’s main customers were Iran Electronics Industry (IEI), which manufactures a diverse range of military goods including missile-related goods and night vision equipment, and Iran Communication Industries (ICI), which manufactures military communications equipment.Footnote 75 Evertop was indicted by the US in 2009 for re-exporting thirty shipments of goods, largely consisting of electronic components, valued at over $1.18 million.Footnote 76 Analysis of the products procured by Evertop, and the means used – including listing a freight forwarder as consignee – suggest that the scheme was an opportunistic effort to procure lower grade goods from naïve suppliers, rather than a sophisticated effort to target companies with significant Internal Compliance Programs.Footnote 77
The activities of Evertop provide some insights into the choice of Malaysia. According to the Iranian nationals running Evertop, the company was a ‘just a small private company stablished [sic] in Malaysia for the sake of shipment purposes only’.Footnote 78 When attempting to procure goods from US companies, they advised them that the end user was in Malaysia.Footnote 79 Apparently – as in the IED case – use of direct Iran Air flights was appealing, being consistently requested from freight forwarders.Footnote 80 Evertop also benefited from the lax regulation of a Malaysian Free Trade Zone (FTZ).Footnote 81 FTZs have frequently featured in proliferation networks presenting a number of vulnerabilities.Footnote 82
In 2008 Kakavand, an Evertop director, sought to establish new intermediate companies. Kakavand listed four generic company names in order of preference for an ‘associate’ to check in the registry.Footnote 83 The nonchalant ‘please prepare the forms until we can sign them’ suggests they faced little difficulty.Footnote 84 Indeed, registry data suggests Evertop was established with no problem by Kakavand and other Iranian procurement agents in 2005.Footnote 85
The second cluster featured a series of companies surrounding Skylife Worldwide Sdn Bhd. A 2009 cable alleges that Skylife and, another front company, Microset Systems Sdn Bhd, allegedly were ‘co-located, work closely with one another, and have acted as brokers for numerous Iranian entities of proliferation concern’.Footnote 86 One of Skylife’s directors Mohammed Mahdavi was described by the US government as ‘an Iranian procurement agent’ and as having worked for SHIG, and Ya Mahdi Industries (an Iranian anti-tank and surface-to-air missile manufacturer).Footnote 87 Microset allegedly worked for a middleman linked to Fan Pardazan and Qods, both entities linked to Iran’s UAV programme.Footnote 88 Both Skylife and Microset also supplied Iranian military aircraft manufacturer HESA.Footnote 89
Enforcement cases provide insights into the operations of this aspect of the network. David Levick, an Australian businessman, allegedly procured goods from US companies and shipped them through a Malaysian company to Iran in 2007 and 2008.Footnote 90 While Skylife is not named in the indictment, reporting suggests that Levick’s first contact with Iranian middlemen was through the company.Footnote 91 The goods included gyroscopes, servo actuators, pressure transducers, an emergency floatation system, and a light assembly for various UAV, aircraft and helicopter applications.Footnote 92 In this case, Iranian middlemen transferred the goods so that they did not enter Malaysian customs territory, with a freight forwarder checking them and re-exporting them in Kuala Lumpur.Footnote 93
In 2008 a complaint was issued against Skylife Director Hossein Ali Khoshnevisrad and his Tehran-based company Ariasa AG.Footnote 94 The complaint included charges that Khoshnevisrad had facilitated three shipments of 17 Rolls Royce helicopter engines from Irish Company Mac Aviation to Iran, via a Malaysian company.Footnote 95 In this case Khoshnevisrad used a separate front company, ‘Pennerbit Kemas Sdn Bhd’, a book trading company according to investigators.Footnote 96 This company appears not to have been registered, and was likely just a false letterhead.Footnote 97 Mac Aviation also transferred other goods to Iran through Malaysia. Aircraft vanes – used in jet engines – were shipped to Kuala Lumpur airport where they were also transferred onto a direct Iran Air flight to Tehran.Footnote 98
By 2009, US officials suggested ‘Malaysia was becoming the “new Dubai” for illicit traders.’Footnote 99 Despite US pressure, the new export control legislation had been ‘floating about in the government, without any domestic champion or political will to push it to fruition’.Footnote 100 In April 2009 Najib Razak replaced Badawi as prime minister, and ‘an unlikely but influential champion for strategic trade controls’.Footnote 101 The Strategic Trade Act (STA), passed in 2010, includes strong penalties for violators such as fines up to $7 million or a death sentence for violations leading to loss of life.Footnote 102
In sum, Iranian procurement activities in Malaysia – for IED, UAV, and missile technology – were extensive. Operations in Malaysia were underway at least a year before UN technology-based sanctions were put in place. However, Iran had been an embargoed destination for US technology since after the Iranian revolution, and subject to various other unilateral restrictions from other developed economies. There is little direct evidence to suggest why Iran chose Malaysia. However, investigation of these cases provides suggestions beyond a basic need to obscure the end user. The lax regulatory environment seems to be appealing, especially around heightened concern and the tightening up of controls in the UAE – Iran’s most significant sanctions-busting hub – in 2007.Footnote 103 Iranian agents could establish new companies to conduct business with relative ease. Again though, besides the lax regulatory environment, present in most of East Asia, other factors may have contributed to the choice of Malaysia specifically. Malaysia, for example, provided the only direct Air Iran link to Tehran in Southeast Asia, which all the Iranian cases above utilised.Footnote 104
North Korean illicit activity in Malaysia
Malaysia has long been a venue for North Korean arms trading, although two cases unearthed in early 2017, likely involving breaches of the UN arms embargo in place since 2006, provide more extensive evidence.Footnote 105 The activities of Glocom, said to be a ‘Malaysia-based company’ advertising ‘radio communications equipment for military and paramilitary organizations’, were featured in a 2017 UN report.Footnote 106 Glocom is described as a ‘front company of the Democratic People’s Republic of Korea company Pan Systems Pyongyang Branch’.Footnote 107 Two Malaysian-registered companies (established in 2005 and 2012) acted on its behalf. The manufacturing of Glocom’s radio equipment appears to have taken place in North Korea, the procurement of components from mainland China and Hong Kong, and export of the finished product conducted using intermediaries based in mainland China, Hong Kong, and Taiwan.
Pan Systems Pyongyang used Malaysia as a ‘base for its key representative abroad’, listing its website at a ‘.com.my’ URL.Footnote 108 Malaysia may also have played a role as a transshipment or transit hub, with a shipment of radio gear being seized on route to a phantom end user in Thailand in 2011.Footnote 109 A 2018 UN report suggested Glocom’s representative in Malaysia disclosed to a Malaysian bank that he intended to set up a factory in the country, although this was likely to justify opening an account.Footnote 110 The network was active recently, with a new Glocom website advertising new products going live in January 2017.Footnote 111 A recent brochure alleged that the company took $10 million annually from transactions in over fifty markets.Footnote 112
A second case involves Kay Marine Sdn Bhd, a Malaysian boat builder, sanctioned by the US State Department in 2016.Footnote 113 A 2011 Kay Marine marketing video featured a number of North Korean-designed military craft including torpedo boats, semi-submersible vessels and a Yono Class miniature submarine. Statements by the company management in 2006 and 2007 suggest collaboration with North Korea in the 2000s, and the relationship may have involved the ‘manufacture of assault boats’.Footnote 114 While there is evidence that Kay Marine marketed North Korean arms, and suggestion of possible collaboration in manufacturing, there is no hard evidence to suggest the vessels were manufactured in or transferred through Malaysia.
Marketing conventional weaponry and related military equipment through entities based in ‘third countries’ is a modus operandi of North Korean arms dealers, allowing them to avoid scrutiny and pass off North Korean military products as goods produced by other countries. The cases involved companies that were to some degree taken over or ‘controlled’, rather than being established, by North Korean agents. North Korea often exploits existing business relationships and historical trading connections. These North Korean cases may not have involved the breach of export controls, because no goods are known to have been transferred through the country, and show how a wider landscape of legislation must be put in place in order to implement UN sanctions.
Why Malaysia? ‘Third country’ selection and proliferation network behaviour
Despite great differences between the cases, they all have at least one common factor: individuals, or groups, decided to exploit Malaysia as a ‘third country’ for proliferation purposes. This is rather than Malaysia featuring by default – for example through its ports due to its position in global transportation networks. This section further explores themes addressed in the cases above – laxly regulated environments and their genesis, as well as other factors such as levels of development, logistical networks, and existing social and political connections. It argues that Malaysia shares the ‘commercial’, but not the ‘geographical’ characteristics, which Early suggests make certain third countries more likely to become trade-based sanctions busting hubs.Footnote 115
Weak regulation and enforcement
A rationale that is evident in all cases discussed is the relative ease of doing business – in terms of export controls, other regulations, and limited oversight. Until 2010, Malaysia did not have a comprehensive export control system in place. Like most countries around the world, it has seemingly never successfully prosecuted a company or individual for breach of export controls.Footnote 116 In the cases explored, the Malaysian government has also displayed a general reluctance to act against proliferators.Footnote 117
While great advances have been made in Malaysia since the SCOPE case, issues remain in Malaysia’s legal framework. For example, in 2015 the Financial Action Task Force noted that, ‘Malaysia’s technical gaps in relation to TFS [targeted financial sanctions] against the financing of proliferation are significant’, with delay transposing new designations a source of concern.Footnote 118 It is unclear whether current Malaysian laws are fully in line with UN North Korea sanctions, and therefore whether the country is equipped to deal with the recent arms marketing cases. Although the STA does cover brokering, it is unclear what legal basis was used by Malaysian authorities to act on the Glocom case.Footnote 119 The 2017 collapse of the Chinpo Shipping trial in Singapore shows the challenges of prosecuting violations when domestic legislation does not exactly reflect the specific language of UN resolutions.Footnote 120
Proliferators have likely exercised ‘jurisdictional arbitrage’ in choosing Malaysia because of its favourable national laws and regulations.Footnote 121 The peril of lax regulation, and the logic of deterrence and displacement when controls are tightened clearly have currency in the US government.Footnote 122 The comment that Malaysia could become the ‘new Dubai’ reflects this logic. Dubai put in place export controls in 2007, around the time the IED smuggling network shifted emphasis to Malaysia. As one analyst from the policy community has noted, Iranian illicit procurement shifted from Dubai to Malaysia as a result of ‘greater scrutiny’.Footnote 123
The ease of establishing or registering companies also constitutes a factor contributing to the attractiveness of the regulatory environment, although evidence in the cases discussed is ambiguous. The Iranian cases seem to have involved front companies being established solely for re-export purposes. However, Iranian networks also involved companies that were not formally registered – merely letterheads.Footnote 124 In the Khan and North Korean cases, companies that had already been established were utilised. In the SCOPE case the company already existed and was renamed. The nature of the relationship between North Korea and Kay Marine is unclear, as is whether the owners of Malaysian surrogate companies of Glocom were aware that North Korean entities were using these businesses. A 2017 World Bank survey suggests that Malaysia is ranked relatively poorly for ease of establishing a business, perhaps suggesting why the Khan and North Korean cases involved already existing companies.Footnote 125
Explaining the regulatory environment: Between economics, politics, and bureaucracy
If a weak or favourable regulatory environment drives proliferators’ choices, what drives countries to develop such environments? Malaysia’s regulatory environment can be viewed as a result of economic, political, and bureaucratic factors. Following a series of structural reforms in the 1970s, Malaysia experienced significant economic growth facilitated by a favourable regulatory environment.Footnote 126 FTZs – established from 1972 onwards, and utilised in some of the Iranian cases above – formed a part of this.Footnote 127 In the 1980s and 1990s Malaysian ports also saw significant growth, competing with other regional players such as Singapore to create the most favourable atmosphere for business.Footnote 128
All cases explored benefited from a Malaysia’s desire for international business. Elements of the Malaysian government have also prioritised economic growth over any political reservations. Government led-efforts to grow Malaysia’s economic relationship with Iran were seen as UN technology-based sanctions were passed in 2006.Footnote 129 In 2007 a Malaysian government Minister suggested Malaysia could help to normalise Iran’s relationship with the international community.Footnote 130 The following year, the country sent a delegation, including firms showcasing potentially sensitive technologies to an Iranian trade show led by official government trade body Malaysia External Trade and Development Corporation (MATRADE).Footnote 131
Around the time that Kay Marine appears to have started a relationship with North Korea – in the mid-2000s – the company’s Managing Director spoke positively about international business.Footnote 132 As recently as 2016 – after the UN’s imposition of sectoral sanctions on coal and iron exports – elements of the Malaysian government still openly pursued a greater trade relationship with North Korea.Footnote 133 In December 2016 the CEO of MATRADE sought to boost ties, stating ‘North Korea is now looking at using Malaysia as a gateway to South-East Asian markets as it finds the country business-friendly with pro-business policies.’Footnote 134
Malaysia’s emphasis of disarmament over nonproliferation, and discomfort with tools such as sanctions and export controls may have also limited enforcement. Malaysia, it was noted in 2008 US government correspondence, while respecting UNSC resolutions, ‘opposes use of sanctions as a means of diplomacy’.Footnote 135 In a statement in April 2004 debates about UNSCR1540, the Malaysian government suggested that the most effective way of preventing WMD terrorism was through nuclear disarmament, and expressed concern about the use of Chapter VII of the UN Charter.Footnote 136
While Malaysia has slowly implemented export control reform, and has shown some reluctance in reining in business ties with Iran and North Korea and acting against their proliferation networks, Malaysia’s relationship with the US has been an important regulating factor. For example, the government Minister speaking of normalising relations with Iran in 2007 caveated it with, ‘I think Malaysia is highly aware of the limits that it can do. We will not do anything to jeopardize our relations with the US.’Footnote 137 However, it was the entry into office of Najib Razak in 2009 that marked the start of a stronger and deeper bilateral relationship with the US.Footnote 138 Razak, as discussed, passed the STA in his first months in office, just before attending the US-hosted 2010 Nuclear Security Summit.
Limited enforcement action could also be explained by the challenges faced by the government bureaucracy in implementing a new export control system. These factors suggest that limited enforcement is a result of inability rather than unwillingness. The Malaysian system has been in place just a few years, for example, and as one former official has noted, has only a handful of staff working on implementation.Footnote 139
In sum, the shape of the regulatory landscape and enforcement efforts are a product of efforts to balance economic development, politics, and security. This balance is sometimes regarded as a zero-sum game – that new efforts to enhance nonproliferation through regulation will impede economic development. Like many developing countries, Malaysia has traditionally been wary that an overly regulated business environment could harm economic development.Footnote 140 The US and others have sought to highlight the positive aspects of creating a more regulated business environment, reducing risk for investors. For example, in 2010 the US highlighted that Malaysia’s ‘legitimate business interests would also suffer if it were seen as a “proliferators’ playground”’.Footnote 141
Explaining network behaviour: Beyond the regulatory environment
Malaysia’s regulatory environment has limited explanatory value alone in explaining the proliferation networks’ choices. Extracts of emails between procurement agents included in US court documents seldom refer to lax regulation.Footnote 142 It is also unlikely that those involved in these networks take such a structured approach to considering relative merits of different jurisdictions.Footnote 143 However, most importantly, the lax regulatory environment does not explain the choice of Malaysia over other relatively unregulated economies in Southeast Asia and beyond. When Malaysia put in place the STA in 2010, only Singapore out of ASEAN’s ten members had put similar legislation in place.Footnote 144 In terms of export controls, by 2010 Malaysia was actually one of the more regulated jurisdictions in the region. Other factors must supplement an explanation centered on a lax export control regulatory environment. These broader factors include features which distinguish ‘hubs’ from ‘havens’ – levels of development as reflected in the workforce, infrastructure and logistical connections – as well as existing social and political connections of those involved.
Basing a hub in a developed and industrialised economy provides a skilled workforce and helps proliferators to import required technology. In the SCOPE case, Malaysia was viewed as having a sufficiently advanced workforce when Dubai and Turkey couldn’t deliver. Kay Marine clearly had something to offer North Korean arms dealers, possibly in technical or procurement terms during alleged collaboration on assault boats, or in terms of an untainted and industrialising economy that could feasibly manufacture and therefore market these vessels without raising suspicion.
Orders for high specification products from a country with limited high-technology industry could raise suspicions about end uses and concern of possible transshipment risk among export controllers in government and industry in advanced economies. This factor likely was a consideration in the Khan and Iranian cases, which involved imports. States with previous nuclear aspirations could also raise concerns. For example, Khan and associates initially discounted South Africa as an option because exports to the country might have raised the interest of intelligence agencies. In the North Korean cases, Malaysia would allow for the goods to masquerade as Malaysian. A less developed country could have been used to market North Korean military vessels and communications equipment, but this could have raised concern about quality, or because the product line was inconsistent with potential buyers’ perceptions of the capability of the country’s industry.
Good logistical and transportation links are clearly important in proliferator’s location choices. Malaysia is a transshipment hub for legitimate global trade, like many other ‘third countries’ exploited by proliferators – for example, Dubai and the UAE, Hong Kong, Singapore, and China. Malaysia is a part of major liner maritime shipping networks and has a large hub airport. Evidence of the importance of transportation links is clearer in the Iranian case. The three Iranian networks discussed saw goods re-exported on direct Iran Air flights to Tehran, making interdiction impossible once the goods left Kuala Lumpur International Airport, then Iran Air’s only Southeast Asian destination. Factors such as access to infrastructure and the costs and time to trade across borders have featured – alongside other factors relating to regulation, compliance, and legal protections – in the World Bank’s ‘Ease of Doing Business’ index.Footnote 145 Malaysia has consistently performed relatively well globally, and in the region in this index.Footnote 146
While evidence of personal and political connections is lacking in the Iranian cases, there are clear examples in the Khan network and the North Korean cases. B. S. A. Tahir’s connections – through marriage in 1998, increasing association with the prime minister’s son, and growing business interests in the country made Malaysia an obvious choice. His connections likely opened doors, and to an extent he believed they would protect him. In the North Korean cases, links to the Malaysian establishment – and that establishment interests have affected enforcement – should not be discounted.Footnote 147 Although it is unclear whether money changed hands, there was clear potential for ‘corrupt protection’ of proliferation networks in these cases.Footnote 148
In sum, while there are clear differences between Early’s concept of trade-based sanctions busting and the use of third countries in proliferation networks, there are also similarities in the characteristics of countries emerging as hubs for these activities. Specifically, as in the UAE example used by Early, Malaysia is an ‘open’ economy, with infrastructure to facilitate international trade.Footnote 149 However, Malaysia’s pre-existing trade links with Iran and North Korea are less extensive than UAE’s links to Iran. The Malaysian case explored undermines the importance of geographical proximity, which Early argues was important in the emergence of the UAE as a sanctions-busting hub.Footnote 150 That proliferation networks see the transfer of a small number of high-value shipments, rather than large volumes of trade, mean that geographical proximity and related low transportation costs are not as important in the emergence of third country hubs in proliferation networks. This agrees with arguments made in passing by Early and Naylor.Footnote 151 While geography is not such an important factor in the selection of third country hubs in proliferation networks, trade-based sanctions busting hubs often also see significant proliferation-related trade as well as trade volumes – for example the UAE in the case of Iran and China in the case of North Korea.Footnote 152
Conclusion: Network behaviour and nonproliferation policy
This article has considered the use of ‘third countries’ in proliferation networks to facilitate the transfer of WMD and military technologies. It has proposed a loose typology of ways that third countries are used in these networks – manufacturing, transshipment, and brokering – and illustrated these using detailed case studies. In doing so, it has sought to provide a more nuanced conceptual grounding for discussion of proliferation networks. The use of Malaysian cases has not been to single out the country – in theory any country can be exploited in such a way, and many have been – but to generate further insights into the decision-making of those involved in proliferation networks.
The article has argued that explanations involving weak regulation and limited enforcement action need to be supplemented with other factors – social, political, bureaucratic, logistical, and personal – to fully understand why those in illicit networks make the choices they do. The article has sought to situate this question, and subsequent findings to contribute to the academic literature on proliferation networks and sanctions busting. Notably, it has found that while a different phenomenon to Early’s trade-based sanctions busting, proliferation networks seek out hubs that share many of the characteristics of trade-based sanctions busting hubs. However, beyond this, our conceptual understanding of how illicit networks operate at the transactional level, and their geography is still underdeveloped.Footnote 153 Moving beyond a basic understanding of how illicit actors behave can provide an empirical basis to inform counter-proliferation strategies.
The importance of other factors highlighted in this study suggests that conducting further research into the relationships between policy tools such as export controls and the illicit activities they are designed to counter may prove useful. Considering the question as to whether export controls and targeted sanctions are having the effects intended by policymakers complements new research, which has highlighted the unintended effects of sanctions.Footnote 154 There also may be insights to gain from further exploration of the literature on how legitimate businesses make their location choices. While data are limited, this article suggests it is still possible to extract insights.
Findings relating to the role of weakly regulated environments, and the importance of other factors such as personal connections, both suggest pessimistic outlooks for policy. There will always be spaces with less regulation and oversight than others – despite some significant successes in the implementation of the UNSCR 1540 agenda. This fact, paired with ‘jurisdictional arbitrage’, suggests that proliferation networks will merely be displaced by efforts to improve export controls and other legal tools, rather than eradicated. In this sense, efforts to improve the implementation of supply-side controls could be viewed as akin to a never-ending quest. That said, improvements in national export control systems could clearly increase prospects for detection and prosecution of these networks, as well as contribute to the development of a norm against illicit WMD-related trade.
On the other hand, focusing on the specific context of these networks and what drives their decision-making suggests that efforts to counter illicit networks should be heavily tailored. Although risk-based approaches to outreach are clearly important in prioritisation efforts, these findings highlight the limitations of indexes in considering third country risks.Footnote 155 Policy should be heavily intelligence driven and focused on disrupting illicit activity where prospects for deterrence is limited. More research should be conducted into unilateral means to disrupt these overseas networks for governments inclined to do so.Footnote 156 In the most prominent ‘third countries’ – namely China – progress on export control implementation and enforcement has been slow, and hostage to bilateral diplomatic relations. Both these types of approach – enhancing legal frameworks and further developing the disruptive toolset – are undoubtedly required, and already being undertaken by governments in their efforts to counter proliferation networks.
Acknowledgements
I am grateful for support and comments on earlier drafts received from colleagues at the Project on Managing the Atom at the Harvard Kennedy School’s Belfer Center for Science and International Affairs: Matthew Bunn, Marty Malin, Will Tobey, and Aaron Arnold. I am also grateful for the useful comments from the anonymous reviewers.
Daniel Salisbury is a Research Fellow at the Centre for Science and Security Studies, Department of War Studies, King’s College London. This article was researched and written during his time as a Stanton Nuclear Security Fellow at the Harvard Kennedy School’s Belfer Center for Science and International Affairs.
