On 25 September 2017, the board of UK-based Imagination Technologies, founded in 1985 (and listed on the LSE in 1994), agreed to a take-over by a Palo Alto-based, Cayman Island-registered private equity firm named Canyon Bridge. The UK’s Guardian newspaper described Imagination as a global leader in designing graphics processors found in Smart phones and other electronic devices.
As part of a week-long series on Mapping China’s Ambitions, The Cipher Brief is partnering with Jamie Burnham to explore China’s Ambitions and threat vectors, how it is organizing to win, what a government ecosystem looks like and the impact that International Collaboration will have in the future.
Jamie Burnham, Research Fellow, Harvard Kennedy School’s Belfer Center for Science and International Affairs
Jamie Burnham is a research fellow at the Harvard Kennedy School’s Belfer Center for Science and International Affairs where he is exploring how digital technologies are changing political intelligence and policy-making. As a British diplomat, he served across Africa and the Middle East, with specific interests in weapons technology proliferation and the resilience of fragile states.
The British company, based in Hertfordshire, had been for sale since June after Apple, its largest customer and a major shareholder, had announced that it would stop using its graphics technology on the iPhone. The move reflected the challenges of global competition in the semiconductor industry, and China’s ambition for science and technology dominance.
Canyon Bridge was funded by the Chinese state council. The Chinese state was pursuing its strategic goal of technology acquisition in order to support its industrial base. Imagination Technologies represented a quick route to getting world class GPU technology. For the United Kingdom, the company represented one of two national leaders in semi-conductor intellectual property. The other company, Cambridge-based ARM, had been bought by Japan’s Softbank consortium. Softbank subsequently licensed ARM’s microprocessor technology to the Chinese government through a joint venture with the Hopu Fund, a state-backed investment vehicle. If the transaction appeared commercial, the motivations were doctrinal. It is a vignette which illustrates PRC’s thirst for knowledge. To quench it, China has evolved large-scale institutional capabilities which are changing the nature of global competition.
Acquisition of Imagination’s semi-conductor technology is one skirmish in what the Chinese Communist Party’s General Secretary Xi Jinping has described as a struggle between Chinese socialism and western capitalism and the need to secure a dominant global position. China’s model complements growth of an indigenous innovation eco-system with external acquisition techniques. It is pursuing a large, state-sponsored strategy which – in the words of the best book on the subject – seeks to “circumvent the costs of research, overcome cultural disadvantage and ‘leapfrog’ to the forefront by leveraging the creativity of other nations.” Intelligence communities tend to focus on the illicit techniques (including cyber theft and industrial espionage). The main practices are in fact licit and represent the use of a range of sophisticated contractual mechanisms that have long been part of the global market for ideas.
Semi-conductors are an example. They are an essential component in the sensors and communications systems necessary for artificial intelligence capabilities. Along with advanced materials, cloud computing and robotics, they are viewed as a necessity for achieving a Chinese vision of leadership in Artificial Intelligence and a future that would ‘promote deep integration between AI and other industries, and form a data-driven, man-machine collaborated, cross-border integrated, and collective intelligent economy.’
The Chinese government has developed highly effective domestic innovation eco-systems, which stretch across a range of high value areas of economic and security competition. In AI, for example, small enterprises are supported by government seed capital and heavy investment in digital education. Technology leaders such as Baidu, Alibaba and Tencent collaborate with government to support the growth of smaller companies. China has now established the largest capital market for AI start-ups, the most research papers on AI, supportive data regulation and is developing the most AI talent. There are, however, significant technology gaps in which foreign competitors maintain superiority. These are identified through a deliberate and large-scale intelligence collection apparatus, and companies and universities overseas are targeted for licit and illicit knowledge transfer. The semi-conductors of Imagination Technologies are an instance of licit transfer. Having acquired the company, the Chinese state seeks to move its intellectual property to China in order to support indigenous technology innovation and prosperity.
This is the age of ‘matrix competition’. If the Cold War was fought on the secretive terrain of defence and intelligence, today’s terrain is global infrastructure which supports trade, communications, technology research, natural resources and supply chains. The UK has long regarded its liberal economic traditions as a source of competitive advantage, encouraging open trade, inward investment and development of an advanced science and technology base. On the country’s intellectual power rest capabilities central to citizens’ long-term security, including nuclear deterrence, cyber power and vaccines. However, its open economy presents risk as well as advantage. Licit or illicit technology transfer systems undermine the UK’s value chains and place jobs, tax revenues and security at risk.
The recent SolarWinds attack should be viewed in a broader context than ‘cyber’. It demonstrates that the size of the threat surface, the potency of relatively cheap techniques and the uncomfortable truth that national vulnerabilities extend well beyond traditional government or agency infrastructure. Pervasive technologies such as the internet of things (IOT) will amplify this trend. Understanding and mitigating the challenge posed by China’s technology transfer capabilities is central to the UK’s security. Hostile activity relies on complex networks operating across transnational infrastructure, challenging traditional assumptions about territorial sovereignty. There will continue to be a blurring of boundaries between domestic and international; between domains of intelligence; between peace and war; between truth and disinformation; and between state and non-state actors.
The volume of available data is likely to increase exponentially. The comparative value of traditional intelligence collection techniques may reduce. ‘Secrecy’ cannot be regarded as a determinant of value. Much data of value will be publicly available. Advantage will be conferred on those states which can identify sources of insight, develop the sensors to collect them, draw knowledge from data in a timely, and deliver operational impact. This requires investment in skills and tools, including advanced analytics technologies such as artificial intelligence. Data will proliferate but it will be divided between structured and unstructured; deep web and surface; digitized and undigitized; qualitive and quantitive. Making sense of it – and taking advantage of it – requires dedicated capabilities.
The UK needs to be able to develop information eco-systems that cross institutional and national boundaries. In confronting the cyber threat, GCHQ’s National Cyber Security Centre (NCSC) has demonstrated the efficacy of reaching beyond the strict territorial boundaries of intelligence. International data-sharing (with suitable policy, security and legal protections) must evolve as a norm. Without effective information collection and analytical systems, policymakers are ill-placed to make well evidenced policy and operational judgements.
CHINESE TRANSFER VECTORS
CHINESE TECHNIQUES FOR KNOWLEDGE ACQUISITION
China has evolved institutional capabilities for technology transfer in order to realise its long-term vision of joining the ranks of the advanced economies. The basic outline of the system was established with the origins of the modern state in 1949, but it was Deng Xiaoping in the 1980s who embraced science and technology as a route to meeting ‘new global challenges and competition’. China-based ‘innovation service centres’ and ‘technology transfer centres’ seek to commercialise technology brought in from abroad through international students, hosting of foreign academics, technology outreach centres in government ministries and affiliated NGOs. Science and Technology ‘co-operation societies’ convey talent and ideas to the PRC while state programmes invite overseas specialists to impart knowledge. Scientific literature is monitored to identify new emergent technologies and innovations.
Three broad ‘transfer vectors’ for acquisition of knowledge can be identified. Extra-licit is used to describe practices which are not subject to external scrutiny, and therefore the legality of the transaction is unknowable.
China-based foreign subsidiaries – Competitions (companies, universities) – Conference and colloquia – Direct technology purchases – Enrollments at foreign universities – Investments / acquisitions of companies – Joint research agreements – Loopholes in trade agreements – Patent mining and exploitation – PRC-backed venture capital funds – Startup accelerators and incubators – State-backed investments in overseas research – Tech exchanges, trade-for-tech agreements – Overseas based labs and representative offices
Breach of contract – Computer network exploitation – Copyright infringement – Insider operations – Reverse engineering – Traditional espionage – Willful patent infringement – Violation of non-disclosure agreements
Extra-licit Transfers (organisations)
Document acquisition facilities – Front organisations for PRC offices – Overseas scholar returnee facilities – PRC ministry offices – Recruiting and brokerage websites – Sino-overseas professional associations – Technology transfer centers – Technology transfer forums – Transfer incentive programmes – Overseas-based facilitation companies – Overseas-based student alumni associations – University-linked ‘innovation’ parks
Extra-licit Transfers (personnel)
Professional facilitators – Administrators of returnee parks and tech transfer centres – Diplomats at embassies and consulates (S&T consuls) – Employees of NGOs that front for declared transfer facilities – Manager of ‘talent’ recruitment and other incentive programmes – Scientists, overseas scholars and entrepreneurs – Attendees at overseas exchange programmes – Awardees of sponsored transfer incentive programmes – Members of overseas Chinese professional associations – Organised alumni of Chinese technical universities – PRC post-docs at overseas research facilities – S&T intelligence workers – Talent spotters attached to factories and labs – Members of S&T ‘business intelligence’ groups – Staffers of China’s open source S&T collection network.
Unpicking PRC activity is challenging. Aspects are highly systematized (patent and standards exploitation, civilian and military open source procurement, ‘talent’ spotting and recruitment) while much – as might be expected in a large and diverse state – is disorganized and duplicative.
CHINESE TECHNOLOGY TARGETING
The Chinese state has evolved sophisticated systems for identifying foreign technology that will promote its national interest. The most comprehensive analysis of these networks is by Hannas, Mulvenon and Puglisi. China has developed national science and technology research system that has, from the outset, combined an intelligence function with conventional information research (the Chinese term 情报 qingbao makes no distinction between the two). A network of over 100,000 science and technology intelligence analysts, supported by advanced analytical systems, conduct state driven research on the latest worldwide science and technology achievements and trends, aiming to reduce the costs and risks associated with innovation. In the 1960s, these analysts were instrumental in supporting nuclear weapons research, satellite and military mainframe computers. This system has now evolved into diverse areas of international competition from quantum computing to biomedicine. They are supported by a network of state research libraries, which aggregate science and technology research holdings. Organisations such as the Institute of Scientific and Technical Information of China (ISTIC) and Military Science Information Research Centre (MSIRC) build and analyse large-scale data-holdings. ISTIC for example holds more than 60 million scientific papers. These centres collect and make available foreign and domestic technology journals materials, manage intelligence reporting, organise intelligence exchanges, perform the analysis and intelligence needed for policymaking, research, and production.
This intelligence analysis system supports a range of international capabilities committed to international knowledge acquisition. Within China, national organisations have been established to support foreign technology transfer. The State Administration of Foreign Experts Affairs (SAFEA), for example, is responsible for identifying foreign specialists who can support domestic innovation. SAFEA has accredited over 80 overseas and 73 domestic organisations which together draw in some 440,000 ‘foreign experts’, per year, a quarter of whom are from western countries and Japan. The Ministry of Science and Technology (MOST) has an end to end responsibility for converting foreign expertise into Chinese capabilities. Its functions include bilateral and multilateral technology relationships and overseas a range of domestic and foreign agencies supporting focused on acquiring foreign technology.
Abroad, diplomatic missions are focused on transferring technology, directly by establishing technology-orientated commercial relationships and indirectly by developing networks of advocacy groups. The functions of the Embassy S&T office are to service China’s ‘domestic S&T plans in priority sectors and project development work.’ Embassy staff engage with advocacy groups to identify areas for support, opportunities for investment and exchanges in China. These groups are open to professionals of any ancestry, but in the US, most are of Chinese ethnicity who have overseas citizenship or have visas to work or study abroad. Groups such as the Chinese Association of Pharmaceutical Scientists or the North American Chinese Overseas Transportation Association include many professionals with advanced degrees. These groups provide a network of specialists which can identify emergent technologies, identify potential Chinese consumer markets and determine how to pass the technology to China in a cost-effective way.
OVERSEAS TECHNOLOGY ACQUISITION INFRASTRUCTURE
China is open about its ambition to secure leadership in high technologies. Its ‘Made in 2025’ industrial strategy paper reflects an economic model that combines centralized planning with elements of private enterprise. It demands that Chinese companies seek intellectual capital from abroad in order to permit their displacement. The intentions are explicit, and the capabilities are organized. State and non-state boundaries are blurred. State activity includes provision of government funding and tax incentives. A range of policy tools place specific demands on foreign companies operating in China to share intellectual property as a cost of business, through joint venture structures, investment zones, special projects, or procurement standards. Legitimate commercial activity also provides a space that permits illicit activity. A high volume of commercial partnerships act as a forum for commercial espionage activity. A 2019 CNBC poll of US companies found that a fifth had intellectual property stolen in the previous year. The value of trade secrets stolen from the US equates to somewhere between $225 billion to $600 billion per year.
Overseas, Chinese companies have used venture capital to target gaps in capability. Acquisitions in semi-conductors, for example, have targeted different elements of the value chain from material to design to manufacturing and testing tools. China has sought emerging capabilities in biotech, health, data analytics, advanced materials, manufacturing, AI and robotics. State-back investment funds research and target emergent technologies and facilitate larger acquisition ploys. For example, Evergrande, a large Chinese property developer, purchased the US company Faraday Future in 2018 which facilitated the likely transfer of electrical vehicle capabilities including R&D, design, production, materials, batteries, sensors and autonomous systems. Often discrete transactions – with no apparent relationship – are in support of a common purpose. CAS’s acquisition of AMD’s and IBM’s server businesses, combined with Hopu’s joint venture with UK’s ARM provide China with the intellectual capital to drive dominance of global computer architecture.
China’s activity can be summarised as having the following characteristics:
- Networked. Many of the capabilities and infrastructure which underpin PRC knowledge acquisition activity comprise social and organisational networks with complex relationships and inter-dependencies. These will often be obfuscated by complex ownership structures. They will be enabled by non-Chinese entities such as law firms.
- Global. Nation states are now merely the terrain upon which complex social networks operate in pursuit of their own interest: a Cayman-registered fund, headquartered in California, seeks the intellectual property of a company in Hertfordshire to support the innovation eco-systems of Beijing. To focus on activity in any one country is to miss the significance of the whole. Licit and illicit activity relies on transnational infrastructure and techniques, challenging traditional assumptions about territorial sovereignty.
- Broad Threat Space. The threat surface extends beyond the narrow confines of the State. PRC networks operate across variable diplomatic, commercial, industrial and academic space. Areas of competition include defence, aerospace, life sciences (including vaccine and therapeutics technology), advanced information technologies and telecommunications, natural resource acquisition and infrastructure. Much of the commercial or academic sectors are oblivious to the potential for longer term harm, while some individuals are complicit. Advice to these communities is vague and ill-informed.
- Data Rich. In the modern world, entities cannot operate without leaving traces in data. Corporate records; satellite imagery; official gazettes; print, broadcasting and social media; flight, shipping and container movements; customs records; property registries offer the opportunity to establish the presence of, and relationships between, individuals and organisations. Relevant data sets will not be confined to any one territory: activity in one country may be reflected in the data of another. For example, exports of a company operating in African country might be visible in the import records of a country in Asia. Value cannot be determined by the sensitivity of collection, but on the insight provided and the impact delivered.
Read part two of The Cipher Brief Special Series on China with Harvard Fellow Jamie Burnham in tomorrow’s Cipher Brief.
Read more expert-driven national security insights, analysis and perspective in The Cipher Brief