Artificial intelligence, big data and autonomous systems along the belt and road: towards private security companies with Chinese characteristics?

Peter Layton

Griffith Asia Institute, Griffith University, Brisbane, Australia

ARTICLE HISTORY Received 30 April 2019; Accepted 4 February 2020

KEYWORDS Peoples Liberation (PLA); Private security companies; artificial intelligence; big data; Belt and Road Initiative (BRI); State Owned Enterprise (SOE); urban surveillance

Introduction

The tasks modern private security companies (PSCs) undertake are once again changing. In the 2000s, driven by the demands of America’s Middle Eastern wars, PSCs went into conflict zones to actively support national armed forces, government agencies and aid organisations.1 In the 2010s, driven by the demands of Russian national prestige, PSCs became instruments of the state, directly advancing national interests through irregular warfare operations.2 In the 2020s, driven by the demands of China’s geo-economic Belt and Road Initiative (BRI), PSCs seem set to advance into a new field: the long-term protection of major commercial investments across regions beset by chronic civil unrest, substantial criminality and incipient insurgencies. This

CONTACT Peter Layton p.layton@griffith.edu.au Griffith Asia Institute, Griffith University, Room 1.30, Macrossan Building N16, 170 Kessels Road, Nathan QLD, Brisbane 4111, Australia

© 2020 Informa UK Limited, trading as Taylor & Francis Group

 

new direction appears likely to draw heavily on information technology advances including artificial intelligence, big data, cloud computing, wireless connectivity, autonomous systems and robotics.

The term ‘private security companies’ is a broad one that encompasses small, local firms operating domestically to large international companies offering a range of offshore services including the ability to operate in conflict zones and use armed force. This article focuses towards the latter end of the continuum, albeit in many cases such companies developed out of those small local firms. The term ‘private military companies’ is sometimes used for this harder end of the PSC continuum but in the Chinese case ‘private security companies’ is more common. It is also arguably more accurate given that the BRI spans several dozen nations, each with their own unique security circum- stances. In this, the broadening of conceptions of what the military domain encompasses plays a part. Military matters now incorporate many operations other than war, including in the cyber realm. The differences between peace- time and wartime activities have blurred as the notion of omnipresent security threats and challenges has gained traction.

In this, PSCs are not mercenaries. Structurally, PSCs are normal corporate entities operating in the global marketplace, endeavouring to be profitable and grow their businesses. They offer a range of services, have strategic plans and seek long-term relationships with clients. In contrast, mercenaries focus on the short-term profits individuals could make through threatening or applying violence. Individual not group competence is prized and the services offered are very limited. If mercenaries privatize conflict, PSCs corporatize it.

In the new direction PSCs appear to be evolving towards, the lead role is being played by Chinese PSCs. In turn the development of these PSCs is being driven by China’s BRI. Announced in 2013, the state-led BRI is a one trillion- dollar global connectivity project encompassing infrastructure, energy, finance and trade. It is a vast mercantilist endeavour that aims to grow the Chinese economy through facilitating extensive international trade across Eurasia and Africa. It has grown into being of a size that it now shapes how China engages its near region and the wider world. By mid-2019, 126 coun- tries and 29 international organisations had signed cooperation agreements, more than 3,000 projects were listed as within the BRI umbrella and some 60 percent of these were underway.3 For Chinese President Xi Jinping, the BRI is ‘the project of the century.’4

Some BRI countries are not uncomplicated investment destinations, hav- ing social conflicts, civil unrest and incipient insurgencies. As more Chinese companies invest in these weak states, security concerns are growing. These worries are being reinforced by an increasing number of attacks on Chinese worksites and workers.5

Moreover, there is evidence that BRI itself can worsen security in the countries and regions it crosses. The cost of BRI infrastructure projects can

 

impose substantial debt burdens on host countries, while creating opportu- nities for graft, corruption and elite rent-seeking behaviour. Such actions can deepen existing tensions, creating social instability and anti-Chinese senti- ment, as has happened in Sri Lanka and Pakistan.6 While China would prefer that local governments manage security completely, not all are sufficiently capable and some need Chinese assistance.

An option is to deploy People’s Liberation Army (PLA) elements but this is unappealing. Such a step would create the perception internationally that the BRI was a geo-strategic initiative motivated by great power politics not an economic one aiming to raise everyone’s prosperity as China asserts. Moreover, deploying the PLA would mean China abandoning its longstand- ing policy of military non-interference in other countries.7 Many would per- ceive China as an imperialist power not a trading partner. PSCs appear the answer to this dilemma. Chinese analysts however, consider the ability of current Chinese PSCs to undertake such international operations to be poor, with the sector still immature.8

This article examines the way Chinese PSCs might develop to meet the BRI’s demands, in particular through employing advanced technology such as large-scale urban surveillance networks, big data, artificial intelligence, bio- metrics, drones, robotics and remote support. The article initially assesses the current characteristics of the Chinese PSC industry sector with its well-funded state-owned enterprise customer base, a reliance on ex-military personnel, a preference to stay within the company compounds the PSC is securing and a symbiotic relationship with the state. The second section discusses how advanced technology is being applied to digitise and automate security in China’s Xinjiang province, Pakistan’s ‘safe-city’ program and Zimbabwe’s national facial recognition program. This technology has relevance to the BRI’s security problems and prudently applied might address some current limitations of Chinese PSCs.

The third section combines the understanding of the Chinese PSC sector with the contemporary Chinese security concepts and technology to suggest a new Chinese PSC model and the capabilities it might possess. This new model is likely to build on, and augment, the currently employed security technology but is not without problems and shortcomings as the fourth section discusses. Concerns include: that a one-size fits all approach may be inappropriate given the diversity of BRI countries; that BRI security is not just a matter restricted to the specific BRI infrastructure sites themselves; that the Chinese strategy in calming Xinjiang province is more than simply applying security surveillance technology and is arguably not replicable elsewhere; and that the Chinese Communist Party (CCP) has worries about PSCs becom- ing well-trained, independent, militia-like groups.

The conclusion brings the discussion together to decide that Chinese PSCs embracing advanced security technologies to meet the BRI’s demands appears

 

plausible. A new form of PSC appears likely to emerge that is different to the American or Russian models. China’s new digitalised and automated PSCs may become ‘PSCs with Chinese characteristics.’

Importantly, for ease of discussion in this article the terms ‘government’ and ‘state’ are used in the conventional way that sees a distinction between society, the state and the government. However, China is not a conventional country. Instead the CCP rules in a uniquely idiosyncratic manner well cap- tured by its vision of building ‘socialism with Chinese characteristics.’ When encountering the terms ‘state’ or ‘government’ this refers to the Party and its manifestations.

Chinese PSC industry today

In China, the market for domestic security services is saturated making the new opportunities offered by the BRI attractive. Going international though, is a complex undertaking requiring an adequate sized company and appro- priate capabilities. In 2016, only 20 Chinese PSCs were offering international services. At that time these companies had about 3200 employees offshore; by comparison only 2600 PLA personnel were then deployed offshore on UN peacekeeping missions.9

The potential international market is very large with almost one million Chinese workers offshore, most employed on large official projects.10 Unsurprisingly for some Chinese PSCs, the BRI is the basis of their entire market strategy.11 John Jiang, managing director of Chinese Overseas Security Group, a small consortium operating in six countries, notes that: ‘One Belt, One Road is a lifetime of work for us, in eight years’ time [2025], we want to run a business that can cover 50 to 60 countries, which fits with the ‘One Belt, One Road’ coverage  ’12

Offshore security is a costly business with the principal customers for Chinese PSCs large State Owned Enterprises (SOE). As an example, the China National Petroleum Company earmarks 3–5 percent of its budget for investments in risky countries for security-related activities. In places where security conditions are extremely unsafe, such as in Iraq, budget allocations can approach 20 percent. In 2015, the company spent more than 2 USD billion (USD) on security. In contrast, smaller private, businesses often under- estimate the risks abroad, favour minimising costs and typically allocate only 1 percent of their budget to security.13

In the field, Chinese PSCs generally adopt a similar operating model. As they often lack the necessary language skills, intelligence gathering capabilities and connections to local governments, most only operate within fenced com- pounds. The compound is then defended from external threats by local PSCs, militias or non-Chinese PSCs.14 In Pakistan for example, Chinese personnel act

 

as security managers inside the Chinese workers’ gated compounds, while the Pakistani army and foreign PSCs guard the outside.15

For Chinese SOEs, this model has some advantages in that security is managed by Chinese PSCs rather than by the SOE directly hiring local or foreign PSCs. This outsources any language or cultural issues to the PSC. Moreover, SOEs have more trust in Chinese PSCs being within the company compound as there is a strong belief that they will protect an SOE’s secrets better than the other PSC options. In Africa, SOEs further worry that local PSCs, while lower cost, could become part of the security problem and not the solution, and may become entangled in tribal feuds or local political disputes.16

Chinese PSCs often differentiate themselves from their foreign counterparts by claiming their employees are unarmed but this is a more complicated matter than it may appear. Within China, there are laws that regulate the arming of domestic PSCs but these do not apply offshore where local laws apply. In Iraq for example, the Chinese PSC staff are not allowed to carry guns, albeit workarounds have been devised. Tao Dexi, a contractor with Dingtai Anyuan International Security, notes that: ‘Chinese security companies have always carried out security missions via local teams  [however] under extreme emergencies, Chinese security staff

can borrow guns from the local security staff.’17

The Chinese government now discourages the use of guns by PSCs operating offshore, fretting that such actions could cause a political backlash in the host country against China. The Chinese government is keenly aware of the negative reputational effects that the Blackwater PSC scandal created for the U.S. government in the 2007 Nisour Square shooting in Iraq and wishes to avoid such notoriety.18 Already, several incidents in Africa and Central Asia have led to Chinese managers firing locally acquired weapons at groups of host nation workers.19

The Chinese state has a strong influence on its nation’s PSC develop- ment. With the PSC’s principal customers being SOEs, the state can steer the SOEs to award security contracts as deemed appropriate. At the moment, the state is keen to see Chinese PSCs rather than foreign PSCs gain SOE contracts. HXZA, for example, has gained a near monopoly on security for Cosco Holding and China Shipping Container Lines, China’s two largest SOE shipping groups. A foreign PSC observed that ‘they clearly have very solid relations to the state, considering how loyal their customer base is. And they are not that cheap.’20

There is a strong business referral network in which SOEs and former Chinese police and PLA officials share long-term friendships, common inter- ests and CCP membership. The contracting procedures and the security tender awards can then come to rely more on the personal networks of a PSC’s CEO than on their company’s capabilities.21 The PSCs themselves almost exclusively recruit PLA veterans.22

 

Given these close relationships, Chinese PSCs do not see themselves being hired by foreign clients; they are instead Chinese firms for Chinese customers. In this there is a reciprocal assumption that agreements the Chinese state has with host country governments can be used to shield Chinese PSC from trouble if needs be.23

This strong state-PSC relationship is not inviolate. While Chinese PSCs use ex-PLA staff, these individuals are generally new to the PSC industry and relatively inexperienced in managing conflict situations in unstable countries like Sudan and South Sudan.24 In 2014, a Hong-Kong based, Chinese-funded, American-operated private security group titled Frontier Services Group (FSG) emerged and quickly gained numerous BRI project SOE clients. FSG was initially headed by Erik Prince, founder of the Blackwater PSC but is now chaired by Chang Zhenming of CITIC, its biggest investor and an SOE. Prince remains with the company as Deputy Chairman.25 The Chinese state appears interested in pushing its PSCs to become more capable and more competitive by accessing foreign expertise.

The differences between Chinese PSCs and a Western PSC like Blackwater are stark and suggest how much Chinese firms need to mature. Tian Buchou, with almost two decades managing security for Chinese companies in the Middle East and Africa, observed after operating with Blackwater that:

Unlike the Chinese security teams who . . . are like doorkeepers, the multina- tional and multilingual Blackwater guards were well trained and provided personnel and cargo protection services . . . [Blackwater] had a comprehensive operational system covering logistics, weapons, high-technology and even medical support.  More than 80 per cent of Chinese security personnel have

just a basic education  and are directly led by people who are just military

enthusiasts and love playing war games.26

Automating security

In far western China, Xinjiang province, officially titled the Xinjiang Uygur Autonomous Region, is both a BRI cornerstone where three major transit routes intersect and a major centre for trialling new security technologies. These commercial information technology innovations are already spreading along the BRI, particularly into Pakistan and parts of Africa. Security is being digitised and automated through large-scale urban surveillance networks, big data, artificial intelligence, facial recognition, biometrics, ubiquitous GPS tracking and smartphone spyware.

In terms of technology, the core of the large ecosystem of social monitor- ing and control in Xinjiang is the Integrated Joint Operations Platform (IJOP). It is an optimised security variant of a military information network system of systems developed for the PLA by a subsidiary of China Electronics Technology Group Corporation (CETC), a major state-owned military

 

contractor.27 Like the military systems, IJOP electronically gathers, stores and shares digital information.

IJOP integrates data collected from numerous widely deployed human sources and machine sensors to create a near real-time common operating picture that guides security activities.28 The data continually collected by IJOP is further used by being structured into a machine readable database format and then stored to become so-called ‘big data,’ a very large data set able to be analysed computationally to uncover patterns, trends and relationships. The big data is then analysed using artificial intelligence algorithms so that IJOP can provide automated identification of individual people of interest detected by the machine sensors, and a prediction of their possible near-term activities. As more and more data is collected, the algorithms are constantly refined and their accuracy and reliability improved through using machine-learning techniques. While data is collected on everyone in Xinjiang, the database is apparently optimised for the particular groups and individuals of interest.

The machine sensors that collect the data automatically feed into IJOP are diverse and spread across the physical and digital worlds. Daylight and infrared facial recognition capable CCTVs are dispersed across the urban environment; ‘wifi sniffers’ collect the unique identifying addresses of com- puters, smartphones, and other networked devices; visitor management systems at numerous checkpoints monitor ingoings and outgoings; biometric security systems check individuals, making use of the DNA samples, finger- prints, iris scans, voice prints and blood types collected from all Xinjiang’s citizens aged between 12–65; cyber systems check the use of Virtual Private Networks and encrypted communication tools such as WhatsApp, Viber and Telegram; speech recognition software screens landline and mobile tele- phone calls; walking gait scanners supplement facial recognition systems; the obligatory Clean Net Guard app on the Uyghur ethnic groups’ smart- phones monitors what is said, read, and written and who they connect with; commercial and ‘spybird’ drones provide video aerial surveillance; GPS track- ers installed in private vehicles track movements; and individuals’ electricity and petrol usage is scrutinized.29

Using this sensor array, the IJOP autonomously tracks the path of indivi- duals across both space and time, looking for changes to the normal pattern of life. When an irregularity or deviation from what IJOP has earlier collected and stored in its database occurs, the system flags this micro-clue to the security authorities as suspicious, prompting an investigation. These devia- tions can include individuals contacting persons of concern, whether locally or internationally, or material changes like activating a new mobile phone. IJOP effectively erects a series of virtual fences that limits system-specified individuals’ freedom of movement.30

The security services in the field interact with IJOP through a CTEC-devel- oped app on their Android smartphones. The app collects and sends back to

 

the IJOP Face++ facial recognition information and local area wireless net- work data. The app is also used in reverse by the IJOP to assign tasks to the security personnel, monitor their performance and score their success in fulfilling tasks. IJOP aims to not just guide the security personnel but also control them ensuring duties are effectively and efficiently carried out.31

The IJOP is always running in the background across Xinjiang, always learning more through using artificial intelligence, big data and cloud tech- nology. The overall intent is both to make security more effective but also more efficient to reduce operating and personnel costs. In this, commercial companies are involved not just in the development but also the operation of the IJOP and its associated security technologies.

A tender released in mid-2017 for a ‘safe city data centre’ in Shache County, Xinjiang is an example of this. The centre sought was to have five floors, four of them underground, and to include IJOP, a command centre, large video displays, videoconference facilities, wireless networked control systems, computer rooms, cloud storage, and high quality security access measures. The centre was to be acquired through a public-private partner- ship using the ‘Build-Operate-Transfer model’ operating over a ten-year period. The first year was for construction, the remaining nine involved the winning bidder operating and maintaining the centre, with the centre handed over to the local authorities on contract completion.32

The Xinjiang security technologies are now being exported commercially. Perhaps unsurprisingly, the first exports are across the border into what Chinese Premier Li Keqiang considers the BRI’s flagship project: the 63 USD billion China-Pakistan Economic Corridor (CPEC).33 Pakistan is no easy country for the BRI to pass through, given its chronic civil unrest and endemic societal turbulence. China’s leaders are though publicly committed to the BRI and have spent considerable political capital advocating for it. China is to a certain degree trapped given Pakistan’s centrality to the project, making security within the CPEC crucial.

China is supporting numerous so-called ‘safe city’ projects to improve security within Pakistan’s major urban centres by providing soft, low interest rate loans and urging Chinese SOEs to become involved. Safe city projects effectively install much of the Xinjiang digital security system with its dis- persed machine sensors across Pakistani cities, including Islamabad, Peshawar, Lahore, Quetta, Karachi, and Gwadar.

Huawei has been involved in building the safe city system in Lahore. This has involved establishing a large, 7000 sqm Safe-City Integrated Command & Communication Centre that uses an IJOP-like system incorporating artificial intelligence, big data and cloud technology. The Lahore safe city system uses some 8000 high-grade CCTV cameras, 4G wireless connectivity, facial recogni- tion, automated vehicle number plate recognition, multiple tracking options,

 

integrated communication platforms, geographic information systems and specialised apps on security personnel’s smartphones.34

Along with digitising Pakistani city security, Chinese SOEs are deeply involved in improving digital connectivity between the two countries, pri- marily through a cross-border fibre-optic cable link. Internet traffic is routed through China with discussion now focussed on adopting Chinese internet regulations and installing a Chinese-style national internet firewall.35 Such connectivity effectively makes Pakistan an outlier of the Chinese network.

In Africa, the technologies pioneered in Xinjiang are also finding custo- mers. In 2018, CloudWalk, a Guangzhou-based tech startup that has received considerable state finance, agreed a strategic cooperation framework agree- ment with Zimbabwe to build a national facial recognition program. The agreement as part of the BRI will see the technology primarily used in security and law enforcement.36 Movement through Zimbabwe’s airports, railways, and bus stations will be monitored using a facial recognition database integrated with biometrics.37

The CloudWalk agreement is the first Chinese artificial intelligence project in Africa and, as such, the technology will need some modification for local conditions.38 The facial recognition database used in Xinjiang is extensive but inappropriate for an African population. Machine learning using a large local database is essential to optimise the Chinese artificial intelligence software for the Zimbabwean urban environment. Accordingly, as part of the agree- ment Zimbabwe will send the facial data on millions of its citizens as captured by CCTV cameras to CloudWalk so it can improve the artificial intelligence systems ability to distinguish between dark-skinned individuals.39

In this, the Zimbabwean database is not the only machine learning option for China’s artificial intelligence systems engineers. In 2017, Chinese company Transsion became the dominant company in the African mobile smartphone market, mainly through optimizing its technology for the African consumer. The smartphones support multiple African languages, dual SIM functionality and cameras that calibrate their exposure settings for darker skin tones, allowing face details to be better captured. This latter feature is important for the company’s new Tecno Camon X Pro that includes facial recognition technology. Transsion will now collect data on millions of customers, allowing the company to improve its facial recognition capabilities through machine learning.40

The need to collect African-specific images to improve the robustness and reliability of ‘big data’ facial recognition systems is readily apparent. The same issue will doubtless exist in Pakistan where the large scale collection of images by CCTV systems collated with biometrics – fundamental to the safe cities’ projects – will be inevitably feed back into Chinese artificial intelligence machine learning software to improve its performance. The broadband fibre-optic link seamlessly connecting Pakistan and Chinese digital networks should allow such artificial intelligence training to be undertaken easily and quickly.

 

A new model Chinese PSC?

The Chinese government wants Chinese PSCs to expand to meet BRI demand, but such growth is problematic given staffing constraints. There are limited numbers of personnel with suitable expertise available; few have more than five years’ experience in the field and most have only worked in a permissive security environment.41 The impact of the skill shortage can be observed in Frontier Services Group’s assessment of much current Chinese PSC planning for higher risk areas. This often neglects ‘risk assessments, training, emer- gency action plans, medical evacuation, tracking, communication, process and procedures, security design and hardware  ’42 On the other hand, the

Chinese state and SOEs have an interest in making Chinese PSCs succeed. Generous funding may be available with the right business case.

Improvements have been forced upon Chinese PSCs providing security for energy industry SOEs operating in southern Iraq. The difficult security envir- onment combined with the Iraqi government’s requirements for licences and regulatory compliance have led to Chinese PSCs developing higher profes- sional standards. Company capabilities and risk management have been enhanced and new technology introduced including unmanned aerial vehi- cles and tethered blimps.

This so-called ‘Iraq Model’ appears an important benchmark for the entire Chinese PSC industry and may be repeated along the BRI. An important aspect of this model transferable to security activities in other BRI countries is the specific objectives the PSCs involved in outhern Iraq have. These are straightforward and uncomplicated involving focussing on safeguarding per- sonnel and guarding infrastructure. The broad intent is simply to protect SOE infrastructure and, in the event of a crisis, evacuate Chinese personnel.43

The various factors of increasing demand for security, skilled personnel shortages, clearly defined security objectives, and the usual commercial drive for improved productivity and profitability all combine to suggest that Chinese PSCs will aim to exploit advanced security technology. The digitisa- tion and automation of security in Xinjiang offers a starting point to consider how Chinese PSCs operating along the BRI might adopt new technology.

The Xinjiang suite of security technologies is comprehensive, spanning an integrated central command and control platform (IJOP), urban surveillance networks, big data, artificial intelligence, voice and facial recognition, bio- metrics, ubiquitous GPS tracking and smartphone spyware. In broad terms, the suite is best suited for a specific location where the various systems can be established and optimised for the local situation. In the BRI though, this may be more difficult than it seems.

Across the BRI there will be some sites that will be discrete and remote, but many others will be within urban areas and some will be very large facilities like ports and airfields. Moreover, the BRI’s connectivity is reliant on developing and

 

keeping secure nation-wide transportation systems. For example, CPEC is a network of roads, rail, pipelines and fibre-optic cables that crosses Pakistan from the Chinese border to the Indian Ocean, near the Iranian border.

The BRI brings a new emphasis on securing areas rather than fenced company compounds. Such a shift may take some time to be fully under- stood. Frontier Services Group noted that many Chinese SOE customers today have misconceptions about operating in high-risk countries:

It is not as simple as hiring a few guards [. . .] With many of the [belt and road] projects covering thousands of kilometres of road, rail, pipeline or hectares of oil field, building a wall or fence is simply not an option.44

Guarding areas is a difficult task, as it traditionally requires large numbers of people; a luxury Chinese PSCs operating offshore do not have. The BRI requirement for securing areas may not be feasible without employing Xinjiang suite technologies as a baseline.

The suite can be expected to have new technologies grafted-on as they become available. A number of changes and new additions appear useful for Chinese PSCs securing the BRI.

The easiest is simply increasing the numbers of machine sensors in the surveillance network. The Xinjiang system pioneered the use of large num- bers of CCTVs capable of facial recognition and in so doing reduced costs through mass production. Costs continue to plummet allowing dramatic increases in the size of CCTV sensor fields. Singapore now plans to install facial recognition cameras on the 110,000 lampposts across the country and is running trials, with three major Chinese firms interested in bidding.45 PSCs at BRI facilities can increasingly be relatively profligate in their use of CCTVs and, given low replacement costs, much less concerned about CCTV acciden- tal losses and deliberate damage.

CCTV facial recognition requires passing the collected images to a central command and control platform where an artificial intelligence algorithm compares these against an optimised database. For a PSC task, such a database is easier to develop, as it only needs to ensure authorised people are within specific areas rather than trying to identify a single person in large, crowded urban areas. A comprehensive biometric database can be progres- sively created as each person is approved and authorised to have access. Over time the database will become more extensive and also more accurate, with fewer false positives as machine learning occurs.

Further downstream as 5G networks become available they will offer the potential to deploy hundreds, perhaps thousands, of Internet of Things (IoT) sensors across fixed sites. Such sensors might be quite simple in surveilling only a small area within a narrow electromagnetic band, but quantity can make up for quality shortcomings. The future Chinese PSC is likely to deploy a cloud of IoT sensors around a BRI location, all sending information back

 

wirelessly to the central command and control platform for artificial intelli- gence algorithms to process, filter, fuse and display as a near real-time picture of the local area.

Such sensor fields can also send information back from authorised per- sonnel to the command and control platform, allowing the personnel to be monitored and tracked in real time. Fitbit activity trackers and certain smart- phones apps offer such capabilities now that are both robust and low cost.

The Xinjiang system now connects personnel in the field to the IJOP central command and control platform through smartphone apps. Such human- machine teaming is being taken further with the introduction of wearable technology, in particular augmented reality glasses. When wearing such equip- ment, PSC staff could access in near real-time facial identification and vehicle plate information held within the company’s cloud-based database. Augmented reality glasses have been recently introduced in Xinjiang province.46

The most significant technical enhancement to the overall Xinjiang suite is likely to be further applications of artificial intelligence. In general terms, systems like the IJOP are excellent at gathering, storing and sharing informa- tion but less successful at processing and contextualising this information. The networks can be somewhat overwhelmed by the volume, velocity and variety of data being collected, affecting their ability to turn the data quickly into useful intelligence.47 This shortcoming is partially overcome in Xinjiang by the simple expedient of employing large numbers of staff but this is not a practical solution for Chinese PSCs operating across the BRI. Incorporating artificial algorithms in IJOP or equivalent systems can help address these issues through:

● improving the integration and processing of information;

● adding new data fusion techniques to improve situational awareness;

● applying machine learning to remote sensing to enhance change detection;

● using ‘expert systems’ for high-level decision support;

● using machine learning for predictive maintenance and fault forecast- ing; and

● using virtual and augmented reality for simulations and training.48

Developments beyond the Xinjiang suite are likely to be in two main areas: mobile autonomous systems and adding remote support. The first will aim to allow PSCs to provide security over much larger areas, be able to quickly change the areas being surveilled and to be able to focus on a specific location in greater detail when necessary. Mobile systems may be essential when securing large ports and airfields and in guarding transportation and communication networks. The second area for development, remote support, will aim to improve the effectiveness and efficiency of the overall security

 

system. Efficacy becomes significant when it is considered that the BRI is a long-duration project, potentially lasting generations.

A variety of mobile autonomous systems are becoming available and in this case may be categorised as drones, unmanned ground vehicles (UGVs) or robots. ‘Drones’ is a popular term for unmanned air vehicles. They can carry sensors able to feed information both to a local operator and the central command and control platform for analysis and display. The combination of low cost consumer electronic packages and plastic moulded air vehicles have allowed mass production and made small drones widely affordable. China is particularly active in hobbyist and commercial drone manufacture; the SZ DJI Technology Company (trading as DJI) has some 70 percent of the hobbyist drone global market share.49

For PSCs, commercial drones can now carry high-resolution camera sys- tems able to identify individuals at ranges of about 800 metres from a drone flying at an altitude of 100 metres. While hobbyist drones have endurance constraints in being able to fly for only about 30 minutes, larger commercial drones can be connected to the ground via a power cable and have much longer flight time providing extended air surveillance. The connection to the ground can also include a high bandwidth communication cable.50 Depending on requirements such a drone could be replaced by a tethered blimp carrying a similar sensor and communication package.

Free-flying drones can be flown forward to investigate emerging situations with some carrying audio systems to warn intruders. Chinese PSCs prefer not to use weapons and instead rely on local PSCs to use firearms however, armed drones may offer the alternative of letting robotics take over. ISIS deployed DJI hobbyist drones dropping hand grenades in the battle for Mosul.51 These proved not just effective against small vehicles but also psychologically debilitating for those encountering them.52 PSCs could do similarly; if not with explosives, then perhaps delivering non-lethal payloads such as smoke and teargas.

The tasks drones can cost-effectively accomplish is steadily increasing with growing interest recently in cargo drones. The overall impact is that Chinese PSCs at BRI sites could now deploy their own organic airpower that auto- mates some tasks, reduces overall personnel requirements and keeps staff inside protected compounds to a much greater extent.

Drones have proliferated and become affordable but unmanned ground vehicles (UGV) are really only just emerging.53 For some decades security companies have operated small bomb disposal robots but these have been highly specialised and in being tele-operated had a very short radius of action. Now such small robots are being fitted with wireless connections, diverse sensors and are all-terrain, including climbing stairs and exploring tunnels. Even so, with a range of about 500 m they have only limited utility for PSCs.

 

Israel’s recent fielding of some 350 Segev UGVs however indicates what might be possible in the future. These UGVs, based on a Ford F-350 truck, can patrol autonomously along a designated off-road route and, if onboard sensors detect an intruder, inform a remote operator who can assess the images the vehicle sends. The Segev has a machine gun the remote operator can use to defend the vehicle or engage detected incursions. In this, Segevs are used where humans are not expected to be present rather than in crowded urban areas; Segevs are not self-driving cars but pre-programmed robots intended to detect changes on their patrol route.54 Chinese PSCs might find such a vehicle useful, at least in its patrol mode, in being able to monitor large unpopulated areas with little human input. Automation can again reduce staffing requirements allowing fewer people to do more.

A variation on this approach is the use of robots for building or large facility security. An example is the Knightscope K-series robots: human-sized, bullet-shaped 140 kg Autonomous Data Machines that patrol programmed locations, have video and infra-red sensors and advise remote operators of any intruder detections.55 While gradually gaining commercial acceptance, these robots face competition from the sharply declining cost of CCTVs. Large CCTV fields may produce results similar to indoor patrolling robots. The counter-argument is that CCTVs are fixed and may be able to be evaded; an autonomous mobile robot has a roaming physical presence that may deter intruders.56

The digitisation and automation of security that the Xinjiang suite, its enhancements, and mobile systems can provide indicates that PSCs could use technology to reduce the numbers of staff they deploy to trouble spots. Machines monitored and controlled remotely could take the place of person- nel working in high-risk BRI areas. The staff structure profile of Chinese PSCs might change with more China-based staff, fewer personnel working in off- shore locations and a large number of deployed robots. Such a remote manning construct further implies that the skills of the in-field personnel may change with fewer doing security tasks and more undertaking main- tenance tasks, attending to the automated security system across the BRI site and beyond.

Implicit in remote manning is the ability of PSCs working at BRI sites to plug into broadband internet services and connect back to China. The invest- ment in the Pakistan-China cross-border fibre-optic cable link fits this require- ment well. This connectivity, in effectively making Pakistan part of the Chinese network, means PSC staff remaining in China can work electronically on digital security tasks in Pakistan almost seamlessly.

Reinforcing this change in PSC manning might be some new forms of support provided by the Chinese government. There is interest in establish- ing a state security agency to coordinate security for Chinese SOEs operating abroad. It is envisaged that the Ministry of State Security would collect

 

intelligence and provide non-traditional security support to the new agency.-

57 The current thinking sees the new agency coordinating Chinese PSC operations in BRI countries.

Such an agency would liaise with the host country government on the PSCs behalf together with sharing with the involved PSCs relevant Chinese national intelligence and data. The data might not just be tactical-level intelligence, but could include technical support for the digital and auto- mated systems the PSCs are using including facial recognition databases and the latest software enhancements to computerised security systems. This directed state help would make PSC security activities in BRI countries more effective and more efficient.

Issues and shortcomings

A new model Chinese PSC would be more capable than the current approach albeit with some issues and shortcomings. The new model would be most appropriate for fixed sites occupied for an extended period where compli- cated digital systems could be installed and comprehensive databases built- up. In this, the new digitised and automated model is compatible with current Chinese PSCs that have a well-funded SOE customer base, predominantly employ ex-PLA personnel, prefer to stay within a fenced compound and be unarmed, and have a close relationship with the state. Moreover, the new model helps address the difficulty of expanding quickly enough to meet BRI demand by using the available personnel more efficiently and being able to access China-based support.

However, issues may arise in reducing interaction by the PSCs with local people in each BRI site. Automation in general, and the Xinjiang suite in particular, reduces people to objects to be managed. This approach may have shortcomings given the BRI is a generational project extending over decades. Forging local relationships might become more important over time with the current Chinese PSC approach of staying inside fenced compounds gradually becoming less efficacious.

Extending this line of argument highlights that in broad terms China is trying to apply the same PSC model across many different situations. BRI sites in some African states may suffer from criminality and disputes over poor pay for local employees; sites in Central Asia may experience clashes with the local populations over fears that an influx of Chinese workers might affect demographics; sites in South Asia may be attacked by insurgents wanting to embarrass the central government and gain political concessions.58 Having a single dominant PSC approach might gain cost-efficiencies but may not be equally effective across countries. Alessandro Arduino is unsure most Chinese companies would realize this issue, noting that:

 

With the exception of the Chinese oil and gas sector, there is a dangerous lack of interest among Chinese SOEs in understanding security. The fact that Chinese corporations still cannot differentiate anti-Chinese violence from political and criminal activity is an alarming example.59

Moreover, the new model Chinese PSC approach shares a significant short- coming with the existing approach. While BRI sites might be well guarded, they remain dependent on the national security situation in the host country. Several BRI countries have governments that may be forcibly changed and the resultant civil strife could devastate BRI investments. Some 20 USDbn was invested in Chinese projects in Muammar Gadhafi’s Libya; almost all of this was lost when the regime was overthrown.60 The 2011 Arab Spring further revealed that paying money to a host country’s armed forces to provide the outer protective screen of a fenced compound secured internally by Chinese PSCs does not work when the military disband or join a civil war.61

These examples may explain the Chinese state’s current interest in helping existing authoritarian leaders install the Xinjiang suite in their countries to assist them staying in power. The suite is though only half of the Chinese strategy for countering social unrest. The Xinjiang suite has been developed to detect, locate and identify particular individuals allowing them to be detained. In Xinjiang province these individuals then enter a massive state re-education system while their family and local community undergo ‘de- extremification.’ The transformation of Xinjiang society is sought with the state purposefully acting to change the culture, customs, and thoughts of the local population.62 The Xinjiang advanced technology suite alone is not enough.

The Xinjiang suite’s technologies have been devised as a key element of the Chinese state’s social control strategy but on the understanding that China has sizeable material means and massive human resources able to implement the strategy. As an example, in Xinjiang province the security measures taken include establishing some 7,500 ‘convenience police sta- tions.’ The province may be more highly policed per capita than East Germany was before its collapse in 1989.63 It is a security model with Chinese characteristics that is unlikely to be easily replicated elsewhere.

Accordingly, the Chinese state has become more interested in mediating between parties involved in intra-state conflicts in South Asia, the Middle East and Africa. PSCs will not be enough to safeguard the BRI and helping auto- crats stay in power may be problematic suggesting making mediation efforts is a worthwhile addition. The Chinese government is well aware that without adequate stability in key countries along the BRI, it may fail. Even so these efforts have attracted some criticism as being short-sighted. China’s media- tion efforts are mostly focusing on conflict management and stability pre- servation rather than long-term issue resolution.64

 

Through taking such seemingly pragmatic approaches to safeguarding its BRI interests, China is gradually moving away from its principle of non- interference in other state’s domestic political structures. In its own way, deliberately seeding Chinese PSCs across the BRI is one step towards chan- ging the Five Principles of Peaceful Coexistence that today constitute the pillars of China’s foreign policy.65

In that regard, the Chinese state’s development of PSCs may cause another even more fundamental problem. The CCP has historically worried about the private ownership of guns and other weapons – especially when they are in the hands of well-trained, tightly organized, militia-like groups.66 Indeed, the PLA itself is not the Chinese state’s army but rather the Party’s. PSCs, as they expand and spread into dangerous countries, may end up armed, as they are now when protecting merchant ships against pirates. Some extend this, foreseeing the possibility that SOEs could use their large budgets to build- up specialized PSCs with military capabilities that might be useful in some future internal Chinese Communist Party power play.67

To counter such a move, in 2010 the Regulation on the Administration of Security and Guarding Services was introduced that stipulated the state must wholly own or be the controlling shareholder of any domestic armed PSC. In this way, Chinese PSCs permitted to carry weapons are administered and supervised by the state.68 This future would seem to await those PSCs operating offshore as well. While the revolving door for personnel from the PLA into the PSC sector may presently be sufficient to guarantee their loyalty, as the PSCs grow the state is likely to move into some form of formal control.

Conclusion

The BRI demand for security services appears set to push Chinese PSCs towards adopting advanced security and social control technologies. This seems a plausible outcome given the various factors involved but is not certain.

In considering PSCs, Elke Krahmann usefully proposed two types of civil military relationships: republicanism and liberalism. Republicanism stressed the central role of the state. It was not only seen as more efficient and effective in providing security but also through centralisation could control security forces better. In contrast, liberalism stressed a small state sector that used market forces and private companies to both gain greater cost-effi- ciency and control security forces through fragmentation.69 If republicanism saw the nation in arms, liberalism advocated contractorisation.

China is a country where the separation of the public and the private spheres implicit in Krahmann’s model is at best immature.70 Nevertheless, the Chinese PSC approach does tend towards liberalism and this will continue to

 

deepen if the digitisation and automation argued in this paper is embraced. However, China has a choice.

The state could switch to the alternative republicanism approach. The state could take-over the PSC industry; this would be uncomplicated given the Chinese state often has large shareholdings in such companies. The PLA (or the People’s Armed Police) could then take over running PSCs in BRI countries. If today’s PSCs principal customers are SOEs, PSCs becoming SOEs themselves or even branches of the armed services does not seem implausible. Indeed, with the current Chinese Communist Party leaderships’ attraction to statist policies and centralisation, bringing PSCs into the fold may at some stage appear attractive.

Against this, as noted earlier in this article, is that China could quickly become seen internationally as an imperialist power. Using PLA units rather than PSCs would generate significant pushback from BRI countries given many are acutely sensitive to any potential re-emergence of colonialism. There would be a real possibility that such a step could cause the BRI project to collapse. Even without this, such a move would inevitably increase bureau- cratization and drive up service provision costs. Instead, for the foreseeable future, China seems intent on remaining with liberalism with its emphasis on, and enthusiasm for, defence contractorisation.

In this clash between a state-centred republican approach versus a com- mercial-centred liberalism approach, it is useful noting the direction Western PSCs have taken over the last decade. The key difference between Western and Chinese PSCs is as Blackwater owner Eric Prince observed in 2007 at the height of PSC boom during the Iraq War: ‘We’re a private company, and there’s a key word there – private.”71 Chinese companies as noted are instead closely associated with the Chinese state; commercial factors are a secondary rather than the dominating concern.

Over the last decade, many major Western PSC companies have been bought out wholly or partly by private equity firms; these include Blackwater, Triple Canopy, DynCorp, Global Strategies Group and Gardaworld. Today, profitability takes precedence with the preferred strategies for achieving this being diversification into different industrial sectors for example cybersecurity, moving away from boots-on-the-ground to less personnel intensive services such as aviation, and creating new short-term companies for each specific contract.72 The later means companies can ‘rebrand in an instant.’73 In analys- ing the British PSC industry, researchers noted: ‘The industry is marked by a high turnover of companies; businesses are formed and then dissolved a short while later.’74

These various strategies means Western companies are less interested in pursuing costly investments in new high technology equipment. Their con- tracts are often only short term and the ability for the company concerned to disappear if incidents or financial issues require is a useful attribute.75

 

Moreover, Western companies can lower their labor costs by recruiting staff from lower salaried countries such as Nepal.76 Chinese companies conversely prefer Chinese staff but, with these relatively scarce and expensive, investing in productivity-improving high technology to secure long-term contacts with Chinese SOEs makes financial sense. This predilection is reinforced by the inability of most Chinese PSCs to disappear given the Chinese state is a major shareholder.

The Western PSC business model has developed in a different direction to the Chinese PSC one. Moreover, the new Chinese digital technology relies to a great degree on ‘big data’ for its effectiveness; such large-scale data is owned by the state. Western PSCs simply do not have access to such information, especially as concerns the numerous countries they operate in. Chinese PSCs and the Chinese state have a symbiotic relationship quite unlike that Western PSCs have with their parent state in the post-Iraq War era.

The future seems promising for Chinese PSCs despite several matters of concern. In simple business terms, the principal unsettled issue seems to be gaining the funding necessary so the various companies can acquire advanced security and social control technologies. Given the Chinese state’s current policy stance on PSCs and its understanding of their importance to the BRI such additional funding will doubtless be provided, most likely by the SOE customers.

Chinese PSCs can be expected to employ advanced security technologies as the companies grow in number and spread out across the BRI. In so doing, a new form of PSC will gradually emerge that is different to the American or Russian models. Chinese PSCs that are task-optimised to protect SOE com- mercial investments along the BRI will be a unique form. If the Chinese Communist Party seeks ‘socialism with Chinese characteristics,’ so China’s new digitised and automated PSCs may become considered ‘PSCs with Chinese characteristics.’

Notes

1. Singer, Corporate Warriors.

2. Marten, “Russia’s use of semi-state security forces,” 198–99.

3. Robinson, “Southeast Asia gains new leverage.”

4. Zhou, “China’s Belt and Road Forum.”

5. Yang, “Securing China’s Belt and Road Initiative,” 2.

6. Legarda and Nouwens, “Guardians of the Belt and Road,” 7.

7. Ramachandran, “Protecting BRI: China’s Foreign Security Concerns.”

8. Arduino, “China’s Private Army,” 41.

9. Ghiselli, “Market Opportunities and Political Responsibilities,” 9.

10. Chan, “Why a private US military firm is of value to China’s belt and road mission.”

11. Legarda and Nouwens, “Guardians of the Belt and Road,” 14.

 

12. Reuters, “Security firms to cash in protecting China’s ‘New Silk Road’.”

13. Ghiselli, “Market Opportunities and Political Responsibilities,” 8–9.

14. Arduino, “China’s Private Army,” 47–48.

15. Basit, “Terrorizing the CPEC,” 708.

16. Arduino, “China’s Private Army,” 103.

17. Clover, “Chinese private security companies go global.”

18. Yang, “Securing China’s Belt and Road Initiative,” 9–10.

19. Arduino, “China’s Belt and Road Initiative,” 18.

20. See note 17 above.

21. Arduino, “China’s Private Army,” 42.

22. See note 11 above.

23. Arduino, “China’s Private Army,” 47.

24. See note 11 above.

25. Adams, “Blackwater Founder Erik Prince’s New Company Is Operating In Iraq.”

26. See note 10 above.

27. “China: Big Data Fuels Crackdown in Minority Region.”

28. Wang, “China’s Algorithms of Repression.”

29. Byler, “Ghost World,”; Wang, “China’s Algorithms of Repression,”; and Chen,

China takes surveillance to new heights.

30. See note 28 above.

31. See note 28 above.

32. Post translated by Google Translate: 1.65 billion | PPP Construction Project of Ping’an City, Shache County, Xinjiang.

33. Boni, “Protecting the Belt and Road Initiative,” 5.

34. Ibid., 7–8. Sponsored, “Here’s How Punjab Police Integrated Command, Control and Communication Centre is Making Lahore Safer,”; PSCA, “Lahore to Get High Grade Surveillance Equipment Under Safe City Project,”; Huawei, “Huawei Announces Safe City Compact Solution to Protect Citizens in Small and Medium Cities.”

35. Boni, “Protecting the Belt and Road Initiative,” 9–11.

36. Chutel, “China is exporting facial recognition software to Africa.”

37. Byler, “Ghost World.”

38. Hawkins, “Beijing’s Big Brother Tech Needs African Faces.”

39. Pilling, “The fight to control Africa’s digital revolution.”

40. See note 36 above.

41. Arduino, “China’s Belt and Road Initiative,” 9.

42. See note 10 above.

43. Arduino, “Protecting Chinese oil.”

44. See note 10 above.

45. Aravindan and Geddie, “Singapore to test facial recognition on lampposts.”

46. Yang, “Chinese AR start-up develops smart glasses to help police catch suspects”.

47. Layton, Algorithm Warfare, 37–38, 41–42.

48. This list draws on: Kania, Chinese Military Innovation in Artificial Intelligence, 10.

49. Thiercelin, “Beyond Personal Drones.”

50. . Bogdanov, “Smart Face Control.”

51. Rossiter, “Drone Drone Usage by Militant Groups,” 118.

52. Layton, “Commercial drones.”

53. Rossiter, “Bots on the Ground.”

54. Layton, “Robot Wars,” 14–15.

 

55. Joh, “Private Security Robots.”

56. Ghaffary, “Is this robot really going to replace a security guard?”

57. See note 10 above.

58. Gafarov, “Rise of China’s private armies,” 43.

59. Arduino, “China’s Belt and Road Initiative,” 10.

60. Arduino, “China’s Private Army,” 72–73.

61. See note 10 above.

62. Doyon, “‘Counter-Extremism’ in Xinjiang.”

63. See note 28 above.

64. Legarda and Hoffmann, “China as a conflict mediator.”

65. Keast, “Speaker Interview.”

66. See note 5 above.

67. Arduino, “China’s Private Army.”

68. Yang, “Securing China’s Belt and Road Initiative,” 9.

69. Krahmann, States, Citizens and the Privatisation of Security, 66, 72–76.

70. Ghiselli, “Market Opportunities and Political Responsibilities,” 3.

71. Blackwater, 109.

72. Black, “What one company’s vanishing act tells us about the private security industry.”

73. Sean McFate on the US PSC industry quoted in Ibid.

74. Bruun, “Britain’s private military and security industry examined.”

75. Fielding-Smith et al., “A Security Company Cashed In on America’s Wars.”

76. Ibid.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Peter Layton is a Visiting Fellow at the Griffith Asia Institute, Griffith University. He has extensive aviation and defence experience and, for his work at the Pentagon on force structure matters, was awarded the US Secretary of Defense’s Exceptional Public Service Medal. He has a doctorate from the University of New South Wales on grand strategy and has taught on the topic at the Eisenhower College, US National Defense University. For his academic studies, he was awarded a Fellowship to the European University Institute, Fiesole, Italy. His research interests include grand strategy, national security policies particularly relating to middle powers, defence force struc- ture concepts and the impacts of emerging technology. He contributes regularly to the public policy debate on defence and foreign affairs issues and is the author of the book Grand Strategy.

Bibliography

Adams, Rosalind. 2019. “Blackwater Founder Erik Prince’s New Company is Operating in Iraq.” BuzzFeed News. https://www.buzzfeednews.com/article/rosalindadams/ blackwater-erik-prince-frontier-services-group-iraq

 

Aravindan, Aradhana, and John Geddie. 2018. “Singapore to Test Facial Recognition on Lampposts, Stoking Privacy Fears.” Reuters: World News. https://www.reuters. com/article/us-singapore-surveillance/singapore-to-test-facial-recognition-on- lampposts-stoking-privacy-fears-idUSKBN1HK0RV

Arduino, Alessandro. 2017. “China’s Belt and Road Initiative: The Evolution of Chinese Private Security Companies.” RSIS Working Paper Series. Singapore: S. Rajaratnam School of International Studies.

Arduino, Alessandro. 2018. “China’s Private Army: Protecting the New Silk Road.” The Diplomat, 20 March. https://thediplomat.com/2018/03/chinas-private-army-protect ing-the-new-silk-road/

Arduino, Alessandro. 2018. “Protecting Chinese Oil: A New Breed of Private Security Company Expands in Iraq.” The Arab Weekly, August 5. https://thearabweekly.com/ protecting-chinese-oil-new-breed-private-security-company-expands-iraq

“Autonomous Security.” 2019. Knightscope. https://www.knightscope.com/

Basit, Saira H. “Terrorizing the CPEC: Managing Transnational Militancy in China– Pakistan Relations.” The Pacific Review 32, no. 4 (2019): 694–724. doi:10.1080/

09512748.2018.1516694.

Blackwater USA. “Hearing Before The Committee On Oversight And Government Reform House Of Representatives.” One Hundred Tenth Congress, First Session, October 2, 2007, Serial No. 110–89, Washington: U.S. Government Printing Office, 2008.

Bogdanov, Vik. 2019 “Smart Face Control: How Facial Recognition Changes Different Industries.” Becoming Human: Artificial Intelligence Magazine. https://becominghu man.ai/smart-face-control-how-facial-recognition-changes-different-industries- f03204496eaf

Boni, Filippo. “Protecting the Belt and Road Initiative: China’s Cooperation with Pakistan to Secure CPEC.” Asia Policy 14, no. 2 (2019): 5–12. doi:10.1353/ asp.2019.0024.

Bruun, Laura, Iain Overton, and Elisa Benevilli. 2018. “Britain’s Private Military and Security Industry Examined.” Action on Armed Violence, December 20. https://aoav. org.uk/2018/britains-private-military-and-security-industry-examined/

Byler, Darren. 2018. “Ghost World.” Logic. https://logicmag.io/07-ghost-world/ Chan, Minnie. 2018. “Why a Private US Military Firm Is of Value to China’s Belt and Road

Mission.” South China Morning Post, July 15. https://www.scmp.com/news/china/ diplomacy-defence/article/2155353/danger-zone-why-private-us-military-firm- value-chinas

“China: Big Data Fuels Crackdown in Minority Region.” 2018. Human Rights Watch, February 26. https://www.hrw.org/news/2018/02/26/china-big-data-fuels-crack down-minority-region

Chutel, Lynsey. 2018. “China Is Exporting Facial Recognition Software to Africa, Expanding Its Vast Database.” Quartz Africa. https://qz.com/africa/1287675/china- is-exporting-facial-recognition-to-africa-ensuring-ai-dominance-through-diversity/

Clover, Charles. 2017. “Chinese Private Security Companies Go Global.” Financial Times, February 27. https://www.ft.com/content/2a1ce1c8-fa7c-11e6-9516- 2d969e0d3b65

Crofton, Black. “What One Company’s Vanishing Act Tells Us about the Private Security Industry.” The Bureau of Investigative Journalism, January 30 (2019). https://www. thebureauinvestigates.com/blog/2019-01-30/private-security-industry

Doyon, Jerome. 2019. “‘Counter-extremism’ in Xinjiang: Understanding China’s Community-Focused Counter-Terrorism Tactics.” War On The Rocks, January 14.

 

https://warontherocks.com/2019/01/counter-extremism-in-xinjiang-understand ing-chinas-community-focused-counter-terrorism-tactics/

Fielding-Smith, Abigail, and Crofton Black, and The Bureau of Investigative Journalism. 2019. “A Security Company Cashed in on America’s Wars—And Then Disappeared”, The Atlantic, January 29. https://www.theatlantic.com/international/archive/2019/ 01/afghanistan-civilian-private-security/581263/

Gafarov, Odil. 2019. “Rise of China’s Private Armies.” Chatham House: The World Today, February & March. 41–43. https://www.chathamhouse.org/publications/twt/rise- china-s-private-armies

Ghaffary, Shirin. 2018. “Is This Robot Really Going to Replace a Security Guard?” Vox. https://www.vox.com/2018/10/8/17913420/security-robot-cobalt-robotics-knight scope-slack-yelp

Ghiselli, Andrea. “Market Opportunities and Political Responsibilities: The Difficult Development of Chinese Private Security Companies Abroad.” Armed Forces & Society 20, no. 10 (2018): 1–21.

Hawkins, Amy. 2018. “Beijing’s Big Brother Tech Needs African Faces.” FP: Foreign Policy, July 24. https://foreignpolicy.com/2018/07/24/beijings-big-brother-tech- needs-african-faces/

Huawei. 2018. “Huawei Announces Safe City Compact Solution to Protect Citizens in Small and Medium Cities.” Huawei Technologies, October 11. https://e.huawei.com/ us/news/smart-cities/201810150942

Joh, Elizabeth E. “Private Security Robots, Artificial Intelligence, and Deadly Force.” UC Davis Law Review 51, no. 2 (2017): 569–587.

Kania, Elsa B. “Chinese Military Innovation in Artificial Intelligence.” June 7 2019 Testimony before the U.S.-China Economic and Security Review Commission Hearing on Trade, Technology, and Military-Civil Fusion. Washington: Center for a New American Security, 2019.

Keast, Jacinta. 2018. “Speaker Interview – Dr. Alessandro Arduino, Codirector of the Security and Crisis Management Program, Shanghai Academy of Social Science.” YCW: Young China Watchers, March 24. http://www.youngchinawatchers.com/ speaker-interview-dr-alessandro-arduino-co-director-of-the-security-and-crisis- management-program-shanghai-academy-of-social-science/.

Krahmann, Elke. States, Citizens and the Privatisation of Security. New York: Cambridge University Press, 2010.

Layton, Peter. 2017. “Commercial Drones: Privatising Air Power.” Lowy Institute: The Interpreter, September 27. http://lowyinstitute.com.au/the-interpreter/commercial- drones-privitising-air-power

Layton, Peter. Algorithm Warfare: Applying Artificial Intelligence to Warfighting.

Canberra: Air Power Development Centre, 2018.

Layton, Peter. “Robot Wars: UGVs on the Battlefield.” Defence Today 12, no. 5 (2018): 13–17.

Legarda, Helena, and Marie L. Hoffmann. 2018. “China as a Conflict Mediator: Maintaining Stability along the Belt and Road.” MERICS: Mercator Institute for China Studies, August 22. https://www.merics.org/en/china-mapping/china-con flict-mediator

Legarda, Helena, and Meia Nouwens. “Guardians of the Belt and Road: The Internationalization of China’s Private Security Companies.” In MERICS China Monitor, edited by Claudia Wessling and Ruth Kirchner. Berlin: MERICS: Mercator Institute for China Studies, (2018): 7, 11–12. Accessed August 6, 2018, 1–18.

 

Marten, Kimberly. “Russia’s Use of Semi-State Security Forces: The Case of the Wagner Group.” Post-Soviet Affairs 35, no. 3 (2019): 181–204. doi:10.1080/ 1060586X.2019.1591142.

Pilling, David. 2019. “The Fight to Control Africa’s Digital Revolution.” Financial Times, June 19. https://www.ft.com/content/24b8b7b2-9272-11e9-aea1-2b1d33ac3271 Plekhanov, Ilya. 2017. “Soldiers without Guns: How Do China PMC.” Center For Strategic

Assessment And Forecasts, July 17. http://csef.ru/en/oborona-i-bezopasnost/340/ soldaty-bez-oruzhiya-kak-dejstvuyut-kitajskie-chvk-7783

PSCA. 2016. “Lahore to Get High Grade Surveillance Equipment under Safe City Project.” Punjab Safe Cities Authority, September 3. https://psca.gop.pk/PSCA/ 2016/09/03/lahore-to-get-high-grade-surveillance-equipment-under-safe-city- project/

Ramachandran, Sudha. 2019. “Protecting BRI: China’s Foreign Security Concerns.” The Central Asia-Caucasus Analyst, March 14. https://www.cacianalyst.org/publications/ analytical-articles/item/13560-protecting-bri-chinas-foreign-security-concerns.html Reuters. 2017. “Security Firms to Cash in Protecting China’s ‘New Silk Road’.” South China Morning Post, April 24. https://scmp.com/news/china/diplomacydefence/arti

cle/2090033/security-firms-cash-protecting-chinas-newsilk-road     Robinson, Gwen. 2019. “Southeast Asia Gains New Leverage as China and US Battle for

Influence.” Nikkei Asian Review, May 29. https://asia.nikkei.com/Spotlight/Cover- Story/Southeast-Asia-gains-new-leverage-as-China-and-US-battle-for-influence

Rossiter, Ash. “Drone Usage by Militant Groups: Explaining Variation in Adoption.”

Defense and Security Analysis 34, no. 2 (2018): 113–126. doi:10.1080/

14751798.2018.1478183.

Rossiter, Ash. “Bots on the Ground: An Impending UGV Revolution in Military Affairs?”

Small Wars & Insurgencies 31, no. 4 (2020).

Singer, P.W. Corporate Warriors: The Rise of the Privatized Military Industry. Ithaca: Cornell University Press, 2003.

Sponsored. 2018. “Here’s How Punjab Police Integrated Command, Control and Communication Centre Is Making Lahore Safer.” ProPakistani, June 19. https:// propakistani.pk/2018/06/19/heres-how-punjab-police-integrated-command-con trol-and-communication-centre-is-making-lahore-safer/

Thiercelin, Arnaud. 2018. “Beyond Personal Drones: How the Largest Drone Manufacturer Transforms Itself.” AI Frontiers, August 11. https://medium.com/aifron tiers/the-evolution-of-drones-how-the-largest-drone-manufacturer-transforms- itself-6b7e7a708f3f

Wang, Maya. 2019. “China’s Algorithms of Repression: Reverse Engineering a Xinjiang Police Mass Surveillance App.” Human Rights Watch, May 1. https://www.hrw.org/ report/2019/05/01/chinas-algorithms-repression/reverse-engineering-xinjiang- police-mass-surveillance

Yang, Yingzhi. 2019. “Chinese AR Start-up Develops Smart Glasses to Help Police Catch Suspects.” South China Morning Post, May 6. https://www.scmp.com/print/tech/ start-ups/article/3008721/chinese-ar-start-develops-smart-glasses-help-police-   catch-suspects

Yang, Zi. “Securing China’s Belt and Road Initiative”. Special Report 436. Washington: United States Institute of Peace, November 2018.

Zhou, Christina. 2019. “China’s Belt and Road Forum Shows Xi Jinping’s ‘Genius Marketing Ploy’ Has Become ‘Project of the Century’.” Australian Broadcasting Commission News, April 25. https://www.abc.net.au/news/2019-04-26/what-is-chi nas-trillion-dollar-belt-and-road-initiative/11042036