IHEDN Mondays are now available in audio format!
Click below to listen:
Mathilde Velliet, who holds a PhD in English Studies from the University of Paris and the University of Aix-Marseille, has been a researcher at the Centre for the Geopolitics of Technology at the French Institute of International Relations (IFRI) since September 2021. Her research focuses on international issues relating to new technologies, in particular US technology policy and Sino-American competition.
Award-winning this year as part of the IHEDN scientific awards, his thesis is entitled «Between security and militarisation: the shaping of US policies on the protection of strategic technologies in the face of China (2009–2025)», and therefore covers the Obama, Trump and Biden administrations.
In your work, you describe how Washington’s objective has shifted from «facilitating high-tech trade» with China to that of «slowing down the adversary». What distinction do you draw between simple security measures (protecting one’s infrastructure) and the build-up of military capabilities (weaponisation) strategic technologies?
Whilst studying the evolution of US policies governing technological exchanges with China, I found that they were based on two rationales.
On the one hand, from the late 2000s onwards, as China began to invest more in the United States and gain market share in certain sectors (such as information and communications technology), the US authorities regarded these Chinese actions as creating vulnerabilities (ransomware, sabotage, cyber-attacks, information manipulation, data harvesting) and sought to mitigate them. The aim was therefore to secure US infrastructure, through a whole series of policies aimed primarily at Chinese «inflows» into the United States, such as imports of 5G equipment or drones, or Chinese investment in strategic US technologies.
However, since the Trump years and even more so under the Biden presidency, these policies to secure infrastructure have been accompanied by measures to slow down Chinese development in certain technologies deemed key, for example by imposing restrictions on «outbound» flows such as sales of advanced semiconductors to China or US venture capital investment in Chinese AI.
Technological interdependencies are therefore viewed both as vulnerabilities to be mitigated and as levers to be «weaponised» – to use the concept developed by political scientists Henry Farrell and Abraham Newman – in order to avoid being technologically outstripped by China. What my work demonstrates is that securitisation and weaponisation are in fact two sides of the same coin: in the context of rivalry with China, protecting critical infrastructure and Measures to safeguard US technological leadership are regarded as essential to national security.
You are examining the period 2009–2025. Can we speak of a cross-party consensus or continuity between the Obama, Trump and Biden administrations? How has the profile of US decision-makers evolved, shifting – to use an expression you quote – from «doves» to «hawks»?
Indeed, whilst the first Trump administration has often been described as a break with the past, or even an anomaly, in terms of technological exchanges with China, I can identify several elements of continuity with its Democratic predecessors and successors.
From the end of the Obama years onwards, the arrival of Xi Jinping and the simultaneous rise of the capabilities and intentions Chinese actions are undermining policymakers« perception of China: as a senior official at the Department of Commerce put it in the 2010s, »we all started out as doves and ended up as hawks’.
In addition to the factors already at play in shaping perceptions of the Chinese threat, under Trump the appointment at various levels of government of figures who are highly hostile towards China is accelerating the «securitisation» process»[1] technological exchanges with that country – and paves the way for the introduction of new, and in some cases unprecedented, restrictions.
As for the Biden administration, it is only revisiting certain measures introduced by Trump in order to correct their shortcomings and make better use of the state’s capacity to weaponise dependencies. Congress – largely overlooked in the academic literature on these subjects – has, across successive administrations, played an active role in framing technological links with China as a security issue.
These elements of continuity reflect the emergence of what I, together with the political scientist Pierre Muller, analyse as a new «global framework» based on new principles:
- a questioning of globalisation, or even a view of international trade as a zero-sum game,
- viewing economic interdependencies as security issues,
- the perception of China as a «near-peer» and the main long-term state-level threat.
Whilst this framework has become increasingly widely accepted since the late 2010s, this does not mean that the political response to this new «common ground» is self-evident! The policies adopted, for their part, are far from being the subject of a consensus, but are the result of negotiations between the various stakeholders in the executive, Congress and even the private sector.
This is the crux of my thesis: to demonstrate that US technology policies are shaped by a combination of the weight of external international constraints (notably China’s rising influence) and the «room for manoeuvre» enjoyed by the actors who articulate these policies and translate them into action.
You show that this policy is not monolithic but is the result of intense bargaining between different actors. How, in practical terms, does the ongoing balancing act between economic prosperity (championed by the Department of Commerce or the Treasury) and national security (championed by the Ministry of Defence or Parliament) play out?
We often hear talk of the reaction of «the United States» or «Washington» towards China, but it seemed to me crucial – building on the work of political scientist Graham Allison and the entire body of literature on Foreign Policy Analysis – to open the «black box» in order to understand which conflicts and compromises between actors led to the US restrictions.
For whilst all the relevant bodies (from the White House to the Department of Commerce, via the Treasury, the Pentagon and Congress) wish to safeguard national security and ensure economic prosperity, not all share the same view of the risks posed by China, or of the appropriate tools and actors needed to address them.
For example, in the interviews I conducted, representatives of the Department of Defence and the «hawks» in Congress who advocate extending the restrictions imposed on China accuse the Department of Commerce of being lax and too close to the business world, whilst the Department of Commerce defends the use of targeted measures and sometimes mocks the Pentagon’s tendency to be «afraid of its own shadow»!
The dividing lines are, however, more nuanced than a strict dichotomy between economic and sovereign ministries: they also pit advocates of an expanded Committee on Foreign Investment in the United States (CFIUS) against advocates of export controls, and those favouring a list-based approach versus a sector-based approach to outward investment, the House of Representatives and the Senate, the Foreign Affairs Committee and the Financial Services Committee… Analysing these «bargaining manoeuvres» is essential to understanding the scope of the measures adopted, their rationale and their chronology.
It should also be noted that this trade-off between prosperity and security – from which the restrictions on trade arise – does not pit two objectives against one another that are as clearly distinct as one might think. It is a long-standing dilemma (one that runs through the history of export controls) whose terms are never set in stone once and for all.
As regards trade with China, this debate reached a turning point in the 1980s, when the US Department of Defence became increasingly reliant on the private sector and its dual-use technological innovations. This development has made the debate more complex, moving beyond a simple dichotomy between security and restrictions on the one hand, and prosperity and the free flow of goods on the other: in some cases, restrictions are detrimental to both security and prosperity, whilst in others the economic benefits generated by the free flow of goods enhance security.
Thus, from the 1990s onwards, a growing coalition (studied by the political scientist Hugo Meijer) argued that liberalising controls also served the interests of security, by enabling American companies to generate revenue through exports, which they could then reinvest in R&D to continue to «outpace» China.
My entire period of study traces the reversal of this economic-security nexus: the very forces that once drove the United States towards openness are now pushing it towards isolation.
You are analysing changes to organisations such as the Bureau of Industry and Security (BIS) within the Department of Commerce. How does the US administration manage to exploit what is known as the bottleneck effect (choke point) to stifle Chinese giants such as Huawei or restrict China’s access to the most advanced semiconductors?
According to Newman and Farrell, states that occupy a central position within the international network structures through which money, goods and information flow can «weaponise interdependencies» through two strategies. The first is the ’panoptic effect«: control over key nodes in the network allows, as in Bentham’s panopticon, a small number of actors to »see« (that is, to gain an informational advantage over) many other actors. The second is the »bottleneck« effect, which makes it possible to cut off certain adversaries» access to the flows circulating through the network.
During the Trump and Biden administrations, the executive branch and Congress have strengthened the capacity of the Department of Commerce – and in particular the Bureau of Industry and Security (BIS) – to exploit these two effects in the context of competition with China. Within the Department of Commerce, the BIS is the office responsible for controlling exports of dual-use goods. Since the end of the Cold War, these controls have been used as a non-proliferation tool, targeting primarily technologies directly linked to weapons development.
I show that, from the late 2010s onwards, they moved beyond this limited use and were gradually employed as a lever in the technological race. Thanks to their central role in certain global supply chains, the United States can exploit the «bottleneck» effect and deprive certain companies – or even the whole of China – of key components, such as semiconductors and the equipment used to manufacture them. Huawei, for example, was placed on a sanctions list in 2019, which deprived it of certain US equipment as well as foreign equipment, thanks to the extraterritorial scope of the restrictions.
Further controls adopted by the BIS in October 2022 prohibit the sale of advanced semiconductors to any Chinese company. The Biden administration has also tasked the BIS with collecting certain data, for example on foreign nationals using US cloud services to train their AI models – thereby illustrating the deployment of the ’panoptic effect«.
I thus draw a parallel between the transformation of the Department of Commerce and that undertaken by the Treasury Department following 9/11: from a secondary player on technological and security issues, it has become the spearhead of the roll-out of technology policies in the service of economic and national security, capable of harnessing interdependencies to respond to the threat.
Washington is pursuing a genuine «diplomacy of exclusion of Chinese technologies». Through which legal and political mechanisms (such as the extraterritorial application of the Foreign Direct Product Rule – FDPR) is the United States managing to compel its allies or third-party companies (such as the Dutch firm ASML) to comply with its restrictions?
This ‘diplomacy of exclusion’ extends domestic protectionist policies beyond national borders, with a consistent twofold aim: to curb allies’ adoption of Chinese technologies deemed unreliable, and to prevent technology transfers – and, under Biden, investment – to China.
With regard to incoming flows, and given partners« openness to Chinese technologies, the aim is to secure the exclusion of suppliers deemed unreliable – first and foremost Huawei in the 5G sector, followed by an expanding range of sectors (semiconductors, aviation, AI). The United States is employing a combination of persuasion, coercion and incentives: »threat awareness« campaigns and the sharing of declassified intelligence to »educate’ allies; threats to downgrade relations or restrict intelligence-sharing should such technologies be adopted; and the promotion of non-Chinese alternatives.
Under Trump, the approach was head-on and largely unilateral. Biden retains this objective but adds a powerful incentive component (funding and development aid conditional on the use of «trusted suppliers») and enhanced cooperation on screening incoming Chinese investment.
When it comes to exports, the Trump and (above all) Biden administrations are attempting to prevent their partners (Japan, the Netherlands, Germany) from selling certain technologies to China, particularly in the field of semiconductors. To this end, they are deploying a range of more or less coercive measures – «a process in which we have cajoled them, held their hand and, at times, twisted their arm», as one diplomat puts it.
The most effective lever is a legal one: the extraterritorial application of US law. Through the rule de minimis [2] and under the Foreign Direct Product Rule (FDPR), Washington subjects goods manufactured to its control regime abroad, which enables it, for example, to block the sale of advanced lithography machines from the Dutch company ASML to China. This legal innovation – achieved through regulatory trial and error under Trump and expanded under Biden – thus represents a major advantage in discussions with other countries, which have no choice but to either have US restrictions applied to their trade, or to implement controls equivalent to those of the United States.
In conclusion, in the face of this systemic Sino-American rivalry, in which technology is used to «win the 21ste »century’, what are the main strategic dilemmas facing Europe and France? Should we adopt the American model of derisking at the risk of technological subjugation, or is there a third, sovereign path?
My work is not intended to be prescriptive, but it does shed light on this very dilemma. The Sino-American rivalry places Europeans in a unique position. Close to the United States in terms of security and values, yet commercially dependent on China, they do not fully share the American perspective that views Beijing as an existential threat.
Since 2019, the European Commission has described China as a partner, a competitor and a systemic rival, but this characterisation masks a wide range of national perceptions. The fear of trade retaliation – a decisive factor, for example, in Germany’s hesitation over Huawei – weighs heavily on any inclination towards alignment.
The most fundamental dilemma is playing out within Europe itself, both between Member States and between the Member States and the European Union. Economic security issues lie at the intersection of the Union’s economic competences and national prerogatives in the field of security. Strengthening European coordination requires ceding some room for manoeuvre to the Commission, something which France and Germany refuse to do, including at the level of the relevant administrations.
But abandoning this approach exposes each state, acting in isolation, to economic coercion from the major powers. The Dutch experience with semiconductors illustrates this well, as Washington preferred to take the discussion onto a bilateral basis rather than see it raised to the European level. A sovereign approach therefore requires first resolving the internal dilemma; otherwise, Europe will be subject to the logic of choosing sides and risks fragmentation rather than negotiating as equal partners.
The European initiatives of recent weeks, notably the package on technological sovereignty, are a step in the right direction by linking the development of European capabilities to a derisking geared towards Chinese but also American dependencies. The use, in June 2026, of
export control measures temporarily denying all non-US nationals access to Anthropic’s AI models, confirms that we should expect increasing use of these national security tools beyond their traditional scope – and that Europe must urgently prepare for this.
[1] A concept derived from critical security studies (theorised in particular by the political science professor Thierry Balzacq in France) which refers to the process of constructing a subject as a «security problem».
[2] A rule setting the minimum proportion of US-origin components that an item must contain in order to be subject to US export controls.
