Leading Edge:

Key Drivers of Defence

Innovation and the Future
of Operational Advantage
Dr Simona R. Soare and Fabrice Pothier

November 2021

The International Institute for Strategic Studies

Contents

Executive Summary 3
Introduction 5
Defence Innovation: Conceptual Foundations 7
Competing definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Conceptual confusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Variable measures of success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Variable sources, drivers and inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Innovation and operational effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Drivers of Defence Innovation 11
1. Threat and vulnerability perceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2. Political-military-societal support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3. Innovation organisation and governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4. Defence-innovation investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Concluding Remarks 28
Notes 30
Executive Summary

Technological innovation is at the core of inter-state While it is premature to assess the efficiency of the
strategic competition. The capacity of nations to suc- five national defence-innovation efforts discussed here,
cessfully develop, integrate and use emerging and the common traits and challenges hold wider relevance.
disruptive technologies (EDTs) in military applica- Five main takeaways are particularly relevant.
tions is a critical element of and metric in the global
competition for power. If innovation has always Purpose and outcomes
been a feature of military competition, the rapid and Clarity is needed around the nature, scope and structure
encompassing progress in a range of emerging and of the process of defence innovation, including in relation
disruptive technologies and a shifting global balance to the expectations about its strategic outcomes. Political
of power are putting pressure on defence establish- and military leaders do not necessarily adopt the new-
ments across the world to adapt and integrate new est technologies, despite their commercial availability,
technologies. Despite profound differences in strate- or use them in innovative ways that enhance their mili-
gic culture, leading military powers are increasingly tary effectiveness and operational advantage. Equally,
focused on how to preserve or create a leading tech- defence innovation is not limited to the use of the newest
nological edge and an operational advantage against and most advanced technologies. Rather, varying cul-
potential adversaries. tural, strategic and historical contexts shape states’ per-
What are the drivers of defence innovation at this ceptions of how new technologies can be integrated into
time of renewed great-power competition and how do defence and new operational concepts developed.
states prioritise among them? How does defence inno- Building a leading edge in defence is not just about
vation translate into operational advantage? What does the adoption of emerging and disruptive technologies
defence innovation take to be efficient and how can it from the civil sector. Nor does this adoption automati-
be measured as such? Through the empirical analysis cally predetermine desirable strategic outcomes in the
of innovation efforts in five countries – China, France, shape of operational advantage and strategic competi-
Germany, the United Kingdom and the United States – tiveness. Defence innovation implies a deliberate tech-
this paper identifies four main drivers of defence inno- nological, organisational and doctrinal change away
vation and offers a better understanding of how these from the status quo, aimed at responding to a new
countries prioritise among them: strategic environment and achieving strategic advan-
tage and competitiveness, including but not limited to
1. threat and vulnerability perceptions, warfighting. This includes creating as well as avoiding
technological surprise. The inherent adjustment implicit
2. convergence in political, military and societal
in defence innovation includes redesigning and imple-
support for innovation,
menting a new relationship between defence establish-
3. innovation organisation and governance, and
ments and societies, particularly expert communities in
4. investment in innovation. private industry and academia.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 3
Threat and vulnerability assessments Innovation organisation and governance
For great powers like the US and China, threat percep- The US is a benchmark for defence innovation and mili-
tions are a key driver of innovation, whereas in the tary power for China and for France, Germany and
case of middle powers, closing domestic vulnerabilities the UK, albeit for different reasons. The American and
takes precedence in shaping defence-innovation efforts. Chinese innovation systems are the most encompassing
For the US and China, defence innovation is about a and structured among the five nations. The opaqueness
new theory of victory based on the disruptive use of of the Chinese innovation system is neither a proof of lack
technology to deliver a strategic advantage against and of challenges nor a measure of its superior efficiency, but
dominance over great power rivals. For middle powers a reminder of the need to develop new qualitative tech-
like France and the UK, defence innovation is a catalyst niques and tools to assess it.
for the preservation of full spectrum deployable combat The British, French and German innovation systems vary
power, a hedge against the negative demographic and in how structured and embedded in the defence enterprise
budgetary trends impacting defence or a coping mecha- they are. However, they encounter similar challenges in
nism for the loss of ‘combat mass’. Concerns over status rapidly scaling innovation solutions and managing failed
are also prominent, as demonstrated by the UK’s recent innovation. Furthermore, they lack direct channels, plan-
emphasis on ‘global Britain’ and France’s warning of a ning, partners at appropriate levels of political decision-
risk of ‘strategic downgrading’ and emphasis on a bal- making and agency to actively promote their innovation
anced approach between innovation and regulation solutions to facilitate their transition to the defence enter-
that increases rather than hampers European strategic prise or the market.
competitiveness. While also qualifying as a middle Human-talent acquisition, retention and upskilling to
power, Germany differs from the UK and France in that feed defence-innovation systems is a strategic challenge.
it considers defence innovation through the prism of None of the countries analysed here has integrated innova-
norms for the use of new technologies and multilateral tion into military education or strategies to attract new talent,
cooperation and only indirectly focuses on the aspect of and planning for how best to use human resources in the
generating military advantage. defence-innovation chain is still limited. China’s practices of
acquisition-hire and attractive fellowship packages are an
Political, military and societal support for increasingly influential tool for attracting foreign talent.
innovation
Deliberate actions to develop and sustain wide politi- Investment in innovation
cal, military and societal support are important for the Beyond the level of investment in research and develop-
sustainability of defence-innovation efforts, which may ment (R&D), the mix between incremental and disrup-
be affected by political instability and societal opposi- tive innovation, as well as the disaggregation of defence
tion. By virtue of its political regime, ensuring top- R&D investment, are important for a more accurate
down and civil–military congruence of priorities for assessment of the efficiency of innovation investment.
defence innovation is more straightforward in China, Very little information is available about how China pri-
though this does not automatically ensure qualitative oritises investments between incremental and disrup-
political steering of innovation efforts. By contrast, in tive innovation. However, the US invests approximately
democratic regimes, political, military and societal sup- 80% in incremental innovation linked to ongoing pro-
port for defence innovation is built through domestic grammes of record and 20% in disruptive innovation,
and political debate, such as in France, the US and the while 40% of its innovation agencies’ spending is chan-
UK. The lack of convergent political and societal sup- nelled towards academia. Meanwhile, European states
port in Germany is a major stumbling block for its are driven by the need to fill enduring capability gaps,
defence-innovation plans. Cultural, political and soci- thus investing many of their R&D funds in incremental
etal barriers to innovation can be just as important as innovation linked to programmes of record managed in
intra-military and technological ones. partnership with traditional defence primers.

4 The International Institute for Strategic Studies

Introduction

Technological innovation is at the core of inter-state stra- through the creation of asymmetric operational advan-
tegic competition. The capacity of nations to successfully tage against great powers like Russia and China. France
develop, integrate and use emerging and disruptive tech- leads on the European continent with its vision of defence
nologies (EDTs) in military applications is a critical ele- innovation defined by the use of EDTs to preserve com-
ment of and metric in the global competition for power. bat mass and global power projection without sacrificing
The impact that EDTs such as artificial intelligence (AI), troop levels. Germany struggles to reconcile its appetite
big-data analytics, advanced sensor and networking for technology and innovation with the social and political
technologies, space and hypersonic missile technologies, reluctance to support such efforts in the name of military
additive manufacturing, advanced materials, next-gener- advantage. China is actively pursuing defence innova-
ation telecommunication systems, quantum computing tion, too, though in an opaque way. China’s military–civil
(QC), human enhancement and bioengineering could fusion strategy may offer an advantage in the exploita-
have on the future capabilities of armed forces elicits sig- tion of EDTs for defence and commercial purposes, and
nificant policy and research interest. Beijing is investing heavily in catching up with the US
Technological progress in disruptive technologies and in leading in areas such as AI and QC.
occurs mainly in the private sector and is driven by com- What are the drivers of defence innovation and how
mercialisation incentives rather than national-security do governments prioritise among them? To what extent
imperatives. This influences the scope and nature of does defence innovation and the development, adoption
ongoing defence-innovation efforts by placing pressure and integration of EDTs translate into operational advan-
on defence establishments across the world. This pres- tage? In beginning to answer these questions, this paper
sure stems from the need to keep up with technological builds on the IISS Defence Innovation Talks series held
progress and the rapid diffusion of advanced technolo- in 2021. It unpacks the concept of defence innovation in
gies to both state and non-state actors, and the challenges relation to EDTs and offers a comparative analysis of five
inherent in adapting commercial technology to military national perspectives from China, France, Germany, the
use. The military fosters a large appetite for technologi- UK and the US. The main contribution of the paper is
cal adoption, but defence structures limit the agility of twofold: to conceptualise defence innovation in a system-
organisational and doctrinal change and engagement atic way and to provide an empirical analysis of the four
with external entities. interrelated main drivers of innovation:
The US is experimenting with new technologies, inno-
vation structures and processes to enable the American 1. threat and vulnerability perceptions,
military to re-establish its operational advantage,
2. convergence in political, military and
including though multi-domain operations. The UK’s
societal support for innovation,
‘Integrated Review of Security, Defence, Development
3. innovation organisation and governance, and
and Foreign Policy’ maps an ambitious plan to integrate
EDTs to displace losses in troop and equipment numbers 4. investment in innovation.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 5
 The paper argues that a well-structured national and seeks clarity in delineating this phenomenon from
system for the long-term management and organisa- related processes, including defence modernisation,
tion of defence innovation is essential and that these military (technological) innovation and the revolution
four drivers are instrumental in gaining a more accu- in military affairs (RMA). The second part identifies
rate understanding of the impact of defence innova- four key drivers that shape defence innovation along-
tion on great-power competition and operational side particular cultural, organisational and contextual
advantage. factors. The final part of the paper offers a summary of
The paper proceeds in three parts. The first part anal- the common aspects of the five national innovation sys-
yses the conceptual foundations of defence innovation tems which hold wider relevance.

6 The International Institute for Strategic Studies

Defence Innovation:
Conceptual Foundations

Defence innovation is often presented in simplified and the concept and whether to focus on its substance or its
ahistorical terms that are either techno-centric, Western-
1
structure. Mahnken, Ross and Cheung define defence
centric or both. Nevertheless, the current wave of defence
2
innovation as ‘the transformation of ideas and knowl-
innovation differs from previous historical instances edge into new or improved products, processes, and services
because of the ubiquity of technology, the process of for military and dual-use applications [emphasis added]’5
globalisation and the diffusion of technological power through collaborations with ‘civil-military science, tech-
from the state to non-state actors (including industry). As nology and industrial base’6 and their application to a dif-
such, this requires a reappraisal of the concept. ferent problem or domain.7 By contrast, Wilkinson and
The availability of technology, human, financial and Jewell define defence innovation as ‘gaining value from the
material resources, the pace of technological develop- exploitation of novelty [emphasis added]’.8 Other scholars
ment, great-power competition and security dilemmas, and practitioners have focused on the development of
capability gaps and the loss of operational capability are innovation pipelines,9 defence-innovation systems, inno-
often cited as drivers of defence innovation. However, vation models,10 strategic innovation agendas, clear suc-
historical evidence suggests a more nuanced approach is cess metrics and risk-tolerant innovation cultures.11
warranted.3 However, the novelty and complexity of EDTs and
Political and military leaders do not necessarily related processes or their provenance from the civil
adopt the newest technologies, despite their commer- sector do not in themselves make it possible to define
cial availability. And, if they do, they do not necessarily
4
attempts to incorporate these technologies into military
use them in innovative ways that enhance their mili- applications as defence innovation.12
tary effectiveness and operational advantage. Equally, The phenomenon of defence innovation implies a
defence innovation is not limited to the use of the new- deliberate technological, organisational and doctrinal
est and most advanced technologies. Rather, varying change away from the status quo, aimed at responding
cultural, strategic and historical contexts shape states’ to a new strategic environment and achieving strate-
perceptions of how new technologies can be integrated gic advantage and competitiveness, including but not
into defence and new operational concepts developed. limited to warfighting. This includes creating as well
It took France considerable time and effort to adopt the as avoiding technological surprise – a concern that is
tank in the early twentieth century, and when it did, its evident in recent narratives about EDTs as presenting
use was guided by old operational concepts which failed both threats and opportunities. Defence innovation is
to provide an operational advantage against Germany in thus the result of the acknowledgement of and rational
World War II. effort to adapt to deep structural macroeconomic (e.g.,
state–industry relations) and changing strategic rela-
Competing definitions tions within and between societies (e.g., hyper-globali-
A first challenge in understanding defence innovation sation-driven interdependence, return of great-power
stems from the lack of consensus on the definition of competition), and to derive power from them. This

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 7
adaptation includes redesigning and implementing a There is wide agreement that innovation occurs during
new relationship between defence establishments and peacetime and is informed by reflections on the previous
societies, particularly expert communities in private conflict and preparations for the next one, whereas adap-
industry and academia. tation occurs in wartime and is geared towards changes
that enhance military efficiency vis-à-vis a specific adver-
Conceptual confusion sary. Robert Work, a former US deputy secretary of
A second challenge in understanding defence innova- defense, made this point, arguing that the requirements
tion stems from conceptual confusion. The definitions of the American operational engagement in Iraq and
discussed above do not accurately capture the differences Afghanistan obstructed a firmer strategic focus on defence
between defence innovation and other similar concepts innovation and related processes of the Third Offset
– notably military (technological) innovation, military Strategy.19 Emmanuel Chiva, the director of the French
adaptation, defence modernisation and RMA. This sec- Agence de l’innovation de défense, has argued that the
tion seeks more clarity between these different concepts, French operational tempo in the Sahel prevents a clear
while also acknowledging that they are interconnected, prioritisation of defence innovation because of urgent
often sequential, and mutually reinforcing phenomena. 13
operational needs.20 While defence innovation concerns
Firstly, defence innovation is often used interchange- warfighting capacity as much as strategic competitive-
ably with military (technological) innovation 14
and, ness with a great-power rival, military adaptation is more
particularly in American circles, with RMA. The three 15
pressingly imposed by operational tempo and involves
concepts refer to a deliberate technological, organisational tactical and doctrinal change more frequently than radical
and doctrinal change away from the status quo and share technological changes, which take longer to develop.21
a techno-centric core. However, military (technologi-
cal) innovation and RMA are driven by defence estab- Variable measures of success
lishments and geared towards warfighting innovation A fourth challenge in understanding the phenomenon of
(capabilities and doctrine). By contrast, defence innova- defence innovation is identifying appropriate metrics of
tion is driven by the more encompassing requirements of success. The International Relations literature is dominated
strategic competition below and above the threshold of by ‘hard’ input factors that define innovation success.22
armed conflict, across the spectrum of hybrid and low- to Such factors include R&D investment; counting national
high- intensity conflict. It is also premised on a civilian research facilities, networks and ecosystems; the number of
and private sector-led process of technological progress. published patents, papers, citations and ongoing projects;
Secondly, the outcome of defence innovation is often the availability of human talent and broad access to foreign
assumed to be the enhancement of military capabilities 16
technology;23 bridging the ‘valley of death’ between devel-
and efficiency. Defence modernisation refers to the incre- opment and procurement;24 and reliable intelligence about
mental improvement and refinement of military platforms the technological efforts of adversaries.25 These metrics are
and equipment, over long-term procurement cycles, to often used as empirical evidence to substantiate public nar-
meet new operational requirements created by the simul- ratives dominated by comparative portrayals of whether
taneous adaptation of adversaries. By contrast, defence- China, Europe or the US is ahead or behind in the integra-
innovation is inherently and purposefully disruptive 17
over tion of EDTs in defence.26
variable time horizons, sometimes overriding existing norms, Less emphasis is placed on ‘soft’ input factors, includ-
procedures and structures. The introduction of the subma-
18
ing strategic culture, organisational and strategic con-
rine, for example, challenged established norms around texts, the quality of the political and military leadership
attack on civilian surface convoys, rendered previously in shaping innovation, threat perception, and qualitative
powerful power projection tools like the dreadnought vul- assessments of innovation outcomes and the national tech-
nerable and changed military doctrine in Europe. nological base.27 This omission partially explains the dif-
Thirdly, conceptual confusion arises from the timing of ficulty of accurately defining when successful innovation
innovation, whether it takes place in peacetime or wartime. has occurred, whether it is process- or product-driven,

8 The International Institute for Strategic Studies

whether change needs to be rapid or incremental in International Relations and Strategic Studies literatures
defence innovation, and why innovation outcomes vary
28
on the sources and drivers of innovation in the mili-
significantly between different international actors with tary. One set of prominent drivers of innovation are
comparable hard-input innovation factors, such as the domestic, including intra-service competition,30 civil-
US and China. Importantly, the measure of success is not ian–military relations,31 known vulnerabilities,32 the
the same for those initiating the innovation and the wider organisation of national innovation systems,33 the dif-
public and expert communities assessing them. Computer fusion of innovation and regulation, strategic culture,34
scientists measure success in their technological innova- the availability of reliable intelligence about adver-
tion process differently than military commanders using sarial innovation efforts,35 and funding. A second set
the algorithm on the battlefield do. Traditional measures of prominent drivers of innovation stem from external
of innovation success refer to crossing the ‘valley of death’ threats36 related to great-power competition and shift-
into capability development and procurement. However, ing global balances of power, including security dilem-
if innovation fails in this, but instead is successfully mas, imperatives of complex deterrence37 and the fear
adapted and adopted by private industry, does this still of war. A third set are horizontal drivers, related to the
qualify it as failed innovation?
29
availability of (new) technology and the pace of techno-
logical progress.
Variable sources, drivers and inhibitors Some of the implicit assumptions of existing theoreti-
Technology is a central consideration for innovation cal approaches – notably, the interdependence between
and warfare alike. The sources of innovation in defence technology and victory or military advantage – are not
are important indicators for measuring its success or fully supported by empirical evidence. The Vietnam War,
failure. However, there is a long-standing debate in the the Iraq War and the recent conflict in Afghanistan are

Figure 1. Summary of main innovation models relevant in defence

Innovation model Main tenants Main proponent(s)
Front-end research/basic research-led
Innovation pipeline Technology push/technology supply-driven model Vannevar Bush
Linear and unidirectional model of progressive nature
Industry-led
Focused on a mix of predominantly incremental innovation (i.e., product development) and some
disruptive innovation Vernon Ruttan
Induced innovation
John Hicks
Technology pull/technology demand-driven model
Bridges ‘valley of death’ challenge
Focused on innovation in production and process
Manufacturing-led William B. Bonvillian
Based on applied R&D integrated in manufacturing process
innovation and Charles Weiss
Industry-led, but industry–government partnerships likely
Focused on both front-end and back-end research
Government-led and funded
Extended pipeline
Mission-oriented, skill-biased innovation model Henry Ergas
innovation
Alleviates crossing the ‘valley of death’ because of deep interconnectedness of government and R&D
institutions
Focused on front- and back-end innovation
Main driver is the strength and capacity of innovation and R&D institutions national innovation systems Richard Nelson,
Innovation
William B. Bonvillian
organisation Improves the means, methods and organisation of innovation efforts and Charles Weiss
Assesses innovation environment to overcome policy barriers to innovation
Innovation management models

Open innovation Agent-led but incorporates external knowledge, innovation, technologies and processes
model Incremental innovation, but open to disruptive external input

Closed innovation Agent-led and confined

model Incremental and sequenced innovation
Source: Compiled by the authors from multiple sources

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 9
examples of technology-savvy great powers prevailing on in tactical nuclear missiles and because of significant
the battlefield with the help, in part, of their superior tech- challenges in its operational use.
nological power, but then failing to achieve a sustainable Innovation requires the replacement of old technolo-
and favourable strategic outcome. This phenomenon is gies, structures and processes with new ones. However,
particularly important in the context of the ongoing great- this neither dictates nor precludes linear outcomes,
power competition where strategic confrontation occurs notably an improvement in military performance, nor
mainly below the threshold of armed conflict and focuses does it mean that innovators will exclusively capital-
on societal cohesion, civilian industry and technology. It is ise on their innovative solutions. Washington derived
also an important reminder that defence innovation does significant operational gains from the introduction of
not by itself produce a favourable strategic outcome. stealth technologies in the 1990s and programmes like
the B-2, F-22 and F-35 progressively refined this capabil-
Innovation and operational effectiveness ity. However, with the rapid proliferation of anti-access,
Finally, the International Relations and Strategic Studies area denial and new generation radar and air-defence
literatures implicitly assume that the main goal of systems, the operational advantage from incremental
defence innovation is to achieve some form of strategic refinements in this stealth technology is declining.
or operational advantage. Yet again, empirical evidence Therefore, in understanding the phenomenon of
does not fully support this assumption. Not all techno- defence innovation, hard as well as soft factors, includ-
logical progress and innovation supports greater and ing context, organisation of innovation and the inte-
sustainable operational and strategic advantage. During gration and use of EDTs, remain equally critical. In
the First World War, a wide range of military innovation addition, the role of cultural and organisational factors
in artillery and in trench and mechanised warfare did in driving or inhibiting innovation is often absent from
not necessarily make war more efficient in that it did mainstream literature on defence innovation. Equally
not deliver military victory against attrition tactics. In important factors impacting innovation and shaping
fact, technological advances do not automatically lead its outcomes are political considerations, including alli-
to leaps in military efficiency and operational advan- ance commitments38 and participation in international
tage when compared to the rate and pattern of adop- organisations and multilateral norm-defining fora;
tion of more disruptive capabilities and the diffusion industrial policy considerations, including protection-
of countermeasures. Despite a rapid R&D process that ism over national defence-industry champions; the
produced a prototype, a successful test and a deployed poor quality of civilian and military leadership, includ-
capability within three or four years, the American M65 ing hubris and organisational (in)flexibility; the pace of
atomic cannon, for example, was obsolete shortly after reform and innovation; and broader societal aspects,
deployment because of more innovative developments including ethical barriers.

10 The International Institute for Strategic Studies

Drivers of Defence Innovation

Drawing on an empirical analysis of efforts in China, efficiency of Chinese innovation as input factors are.
France, Germany, the UK and the US, we identify four Gaining such insights into what happens ‘in the mid-
interrelated defence-innovation drivers in the context of dle’40 of Chinese innovation projects is key to avoiding
renewed great-power competition: threat and vulnerabil- both overestimating the impact of certain investments as
ity perceptions; convergent political, military and societal well as technological surprise.
support for innovation; clear innovation organisation and Expanding wargaming and experimentation is an
governance; and sustained investment in innovation. integral part of the United States’ Defense Innovation
Initiative announced in 2014.41 In 2019, the UK established
1. Threat and vulnerability perceptions a new Defence Wargaming Centre (DWC) within the
The main driver of defence innovation is a consistent, Defence Science and Technology Laboratory (Dstl), one
clear and interdependent threat and vulnerability assess- of the Ministry of Defence’s main innovation foundries,
ment, which, in turn, depends on the capacity to explore and accelerated the number of wargaming exercises.42
and identify threats and vulnerabilities early. In 2019, France established the ‘Red Team’ initiative to
Foresight, wargaming and reliable intelligence are develop foresight and wargaming scenarios that inform
indispensable tools in generating reliable threat and vul- defence-innovation efforts.43 In 2020, the Bundeswehr
nerability assessments as well as supporting the early led the ‘Resilient Response 2020’ multinational wargam-
identification of the emerging security threats and risks 39
ing exercise which enabled the Multinational Medical
that drive defence innovation. It is worthwhile consid- Coordination Centre / European Medical Command
ering whether available intelligence tools and practices (MMCC/EMC) to explore the impact of COVID-19 and
are providing a deep and accurate understanding of the similar pandemic shocks on security and defence.44
defence-innovation efforts of adversaries in ways that go Reportedly, Beijing has also intensified its use of war-
beyond input and output factors. Available data offers gaming in the context of military–civil fusion, includ-
insights into the limits of Chinese reverse engineering ing the large-scale CASIA-Prophet wargaming exercise
and Beijing’s continued reliance on industrial and cyber which focused on AI-enabled Chinese command and
espionage for access to foreign technology and innova- control in the context of the air–sea battle.45
tion. However, it does not offer knowledge about how Other external factors, including alliance participa-
the Chinese government measures progress in innova- tion, ongoing major operations and interoperability
tion, how coordinated Chinese innovation efforts are, requirements, present different pressures in support of
whether Beijing rewards access to external innovation or inhibiting defence innovation. While alliances like
more than fostering domestic innovation, what pressure NATO act as a beltway for technological diffusion and
points Chinese institutions encounter in implementing interoperability standardisation, multilateral approaches
innovation projects, how they are funded, or who the to defence innovation are lacking in competitiveness by
key stakeholders are and how they are related to each comparison to national efforts.46 By virtue of their secu-
other. Yet these factors are just as consequential for the rity guarantees, alliances may also alter threat perception

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 11
and obscure vulnerability assessment, which does not Their mutual perceptions of the other as the long-term
incentivise innovation. Similarly, ongoing operations great-power competitor/threat, and of their status,
can focus military and political leaderships on adapta- prestige and leadership in the international system,
tion requirements in ways that inhibit rather than stim- are the main drivers of American and Chinese defence
ulate defence innovation. While there is little appetite
47
innovation. The United States’ 2018 National Security
for cooperation on specific uses of emerging technolo- Strategy and 2021 Interim Guidance on National
gies in defence among France, Germany, the UK and Security prioritise the shift towards countering the
the US, because of the highly sensitive nature of these long-term challenge of a rising China. The Chinese
technologies, there is nevertheless scope for cooperation challenge has infused and taken centre stage in all
on interoperability requirements, standardisation and US engagements, from its military alliances50 to trade,
norms-building. Middle powers like Germany and the emerging technologies and telecommunications,51
UK regard interoperability requirements as a powerful cyber, space, climate change, nuclear-arms control52
stimulus for innovation, including through the lens of and the Arctic.53 Equally, China’s defence-innovation
the legitimation it offers to innovation efforts and lock- efforts are a continuation of its decades-long defence
ing research and development and innovation (R&D&I) reform and modernisation geared towards catching
funding in place. 48
up with, and eventually surpassing, the US, seen both
Despite a general sense of urgency attached to as a powerful adversary and a benchmark in military
emerging technologies, national defence-innovation power and geopolitical clout.
priorities are highly dependent on contextual, organisa- The defence-innovation efforts of both these great
tional and cultural factors. The prospect of great-power powers aim to develop a new theory of victory, based
competition below and above the threshold of armed on a technophile approach to shaping a new revolu-
conflict features consistently in the threat assessments tion in warfare and altering the balance of military
of the five countries analysed here, albeit in distinct power. The realisation that all great powers were
ways. For great powers like the US and China, defence rapidly achieving military-technological parity in
innovation is about a new theory of victory based on precision-guided munitions and information-driven
the disruptive use of technology to deliver a strategic battle networks, and eroding American technological
advantage against and dominance over great-power superiority and dominance, triggered the launch of the
rivals. For middle powers like France and the UK, 2014 Defense Innovation Initiative.54 The status quo was
defence innovation is a catalyst for the preservation increasingly unsustainable in the context of Chinese
of full-spectrum deployable combat power, a hedge military planning geared towards closing strategic gaps
against the negative demographic and budgetary with the US and achieving military-technological parity,
trends impacting defence, and a coping mechanism if not leadership, in EDTs like AI and QC. Consequently,
for the loss of ‘combat mass’. While also qualifying the US is experimenting with multi-domain operations
as a middle power, Germany differs from the UK and and enhanced information superiority to identify new
France in that it considers defence innovation through sources of strategic advantage on and off the battlefield.
the prism of norms for the use of new technologies and The 2019 White Paper on ‘China’s National Defence
multilateral cooperation, and only indirectly focuses in the New Era’ refers to persistent ‘risks from technol-
on the aspect of generating military advantage. ogy surprise and a growing technological generation
gap’ and posits that a new revolution in military affairs,
Purposeful disruption deemed ‘intelligent warfare’, is ‘on the horizon’.55
and a new theory of victory Beijing’s ambition goes beyond military modernisation
American and Chinese defence-innovation efforts are and towards achieving military-technological leader-
informed by their perception of the shifting global bal- ship, especially in EDTs.56 Though Western analysts
ance of power, their dyadic axis of strategic competi- remain sceptical of Beijing’s track record in transition-
tion and persistent features of their strategic cultures. 49
ing R&D projects into actual capabilities and training,57

12 The International Institute for Strategic Studies

China is exploring ways to become a ‘first mover’58 in integration of key technologies and the operationalisa-
frontier technologies such as AI, QC, biotechnologies tion of the Quad working group on technology.
and telecommunications, and in their applications
in surveillance, early warning, enhanced situational Catalysts and coping mechanisms
awareness, cyber and space, mobilisation and logistics In the case of middle powers such as France, Germany
and information operations. and the UK, specific strategic and cultural contexts
Critically, the Chinese approach to ‘intelligent war- shape defence-innovation patterns both in relation
fare’ leverages emerging technologies to try to create to external-threat perceptions and assessments of
enduring technological gaps in areas not easily acces- domestic vulnerabilities. However, empirical evidence
sible to humankind, 59
especially information, cyber suggests that in the case of middle powers, closing
and space. The focus of Chinese military thinking is
60
domestic vulnerabilities takes precedence in shaping
on its future power and capabilities, imagining a future defence-innovation efforts. Concerns over status are
battlefield in which these technologies could transform also prominent, as demonstrated by the UK’s recent
the speed and conduct of operations, whether through emphasis on ‘global Britain’ and France’s warning of
the use of unmanned systems across all domains or AI a risk of ‘strategic downgrading’66 and emphasis on a
used as an enabler, including of decision advantage. balanced approach between innovation and regulation
The benefits of AI technologies for China are critical for which increase rather than hamper European strategic
improved situational awareness, detection, decision- competitiveness.67
making advantages, targeting and strike capabilities Notably, the drive to retain power-projection capa-
and, thus, may reduce the cost of power projection. 61
bilities, maintain full-spectrum capabilities for low- to
However, for China they also create strategic vulner- high-intensity multi-domain warfare, and remain pre-
abilities, which include higher risks of cyber- and pared for modern deterrence in the context of new bat-
electronic-warfare attacks, lower thresholds of war and tlefield developments68 and limited national material
potential negative impacts on strategic stability, as well resources are the primary considerations for France and
as new ethical issues.62
the UK, albeit their approaches are different. Several
Whereas American strategic culture is still predomi- factors complicate choices for the UK, including the
nantly premised on the belief that military-technological depletion of British capability69 and troop70 numbers,
superiority ensures military dominance63 through tech- inherent vulnerabilities created by capability gaps
nological integration and multi-domain operations, arising from the uncoordinated retirement of legacy
Chinese strategic culture is informed by a sustained military platforms and the fielding of next-generation
effort to catch up militarily and technologically with the capabilities, and ongoing operational needs. The UK’s
US and with its own status as a great power in interna- 2021 Integrated Review presents new technological
tional relations. The technological race with China is not solutions as partial offsets for the accelerated loss of
a conventional arms race but rather ‘a values competi- ‘combat mass’, whereas Paris regards them as enhanc-
tion at its core’, which seeks to define the new rules
64
ing its power-projection and combat capacity. However,
of the road rather than merely regulate the use of tech- combined with the inertia of long-cycle capability
nology. It goes to the core of strategic competitiveness.
65
development, aversion to vulnerabilities created by the
As demonstrated by the calls for US-led technological reduction in legacy platforms is an important inhibitor
alliances, the link between trade and technology, and of defence innovation in France, Germany and the UK.71
the meshing of democracy and strategic technological To fully leverage technological solutions as replace-
goals, alters US views of the role of technology not just ments for loss of warfighting capacity, the UK is exper-
within established alliances, but in shaping and driv- imenting with a new doctrine of warfare based on a
ing new alignment and alliance patterns. This has been novel balance between force concentration, dispersion
recently demonstrated by the AUKUS deal, the eleva- and speed. It also requires a new way of warfare for
tion of the role of the Five Eyes alliance in the military the ‘era of constant competition’.72 Achieving a leading

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 13
edge means either the recapitalisation of combat mass, works to increase France’s overall competitiveness in
in the French case, or using EDTs for asymmetric oper- the context of great-power competition is essential.82
ational advantage and a way of escaping the ‘tyranny In the case of Germany, the main drivers are the need
of symmetry’ in inter-state high-intensity conflict in for wider access to and keeping pace with technologi-
the case of the UK. 73
Both approaches demand new- cal innovation in civilian industry83 and ‘external nudg-
force generation and force-structure models based on ing’ by close NATO allies in relation to interoperability
a mix of open-architecture inhabited and uninhabited requirements.84 Multilateral processes in NATO and the
platforms, including attritable74 ones, that recapitalise EU geared towards defence innovation and the adop-
mass and manoeuvrability in the armed forces. tion of EDTs in defence have a significant shaping and
For example, the UK is entertaining the prospect legitimising power in the German context, even though
that by 2030 a tactical formation in the Royal Air Force they are always linked to long-term sovereign German
‘will move from being eight typhoons, to being two defence planning. New technologies are important for
Typhoons, 10 Mosquitoes and 100 Alvino’, 75
whereas Germany, not least because of interoperability con-
across the air and maritime domains next-generation cerns, negative demographic and budgetary trends and
British military platforms may be by default uncrewed, defence-industrial-policy reasons. Admittedly, defence
with the option to deploy as an inhabited variant. 76
innovation was neither a top-level priority nor a struc-
The French project SCORPION (synergie du contact tured process85 before the launch of the NATO and
renforcée par la polyvalence et l’infovalorisation), the EU innovation processes in 2018–19. Arguably, EDTs,
Future Combat Air System (FCAS) and the UK-led innovation and capabilities are more present today in
TEMPEST project focus on the application of human– German defence narratives than a few years ago, poten-
machine symbiotic and collaborative fighting doctrine. tially indicating a development in the political mindset
France’s project ARTEMIS (architecture de traitement et of the Federal Ministry of Defence,86 though perceptions
d’exploitation massive de l’information multi-sources) persist among the military that close alignment with the
seeks to develop a ‘sovereign infostructure’ for the French US is the best chance for Germany to be innovative in
armed forces, while project ARISTOTLE is developing
77
defence.87 However, programmatic documents like the
an AI-enabled automated operational data-management German White Paper on Defence and the 2021 position
prototype unit. The newly launched French project paper on the future Bundeswehr only vaguely mention
VULCAIN aims to accelerate the integration of mili- the role of EDTs and innovation.88
tary robotics into French tactical land units78 as well In the German political and organisational culture,
as assist with the testing of these new capabilities and EDTs and defence innovation are ‘catch words’ in that
the resulting reorganisation of tactical command. The they are publicly portrayed as necessary for Germany
British Army has already operationally tested its tactical to maintain modern armed forces, but for cultural rea-
AI-enabled data-processing prototype during Exercise sons do not translate into any measure of improve-
Spring Storm in Estonia in June 2021. The Dstl and the 79
ment in German battlefield efficiency. Unlike the other
Defence and Security Accelerator (DASA)’s Intelligent national approaches, German innovation priorities
Ship project seeks to exploit a ‘human–machine network are deliberately modest in scope and level of disrup-
that could work collaboratively on military platforms’ tion. They focus on improved command and control,
by changing the approach of the British armed forces enhanced situational awareness and information dom-
to decision-making, mission planning and automation inance through the enhanced networking between
from 2030 onwards. 80
decision-makers, sensors and capabilities. Deterrence
The accelerated pace of technological progress and and threat perceptions of great-power competition
diffusion to state and non-state actors is a key driver of are admittedly German concerns only in the multi-
innovation for France and Germany. French officials pre- national context of NATO.89 Technological solutions,
sent ‘the democratisation of technology’ as a ‘key trig- including human–machine teaming, are desirable to
ger’ for defence innovation, and ensuring innovation
81
tackle pressing negative demographic and budgetary

14 The International Institute for Strategic Studies

trends. Human–machine teaming is actively explored 2. Political-military-societal support
especially in the context of the FCAS project, though The second factor identified as a key driver of defence
the German preference remains for inhabited military innovation is political, military and societal support.
platforms wherever operationally and technologically Political and institutional contexts shape political–
feasible. 90
However, as demonstrated by the 2020–21 military and societal relations in respect to defence
parliamentary debate which has so far prevented the innovation and consistent political and societal
Bundeswehr from procuring armed Heron TP drones – support for defence innovation is a necessity. The
or indeed autonomous armed variants of the Eurodrone 91
strength of political support and direction of national
– any measure of autonomy in lethal weapons systems innovation efforts varies significantly across the five
is strongly rejected by the German political spectrum national perspectives analysed here. This has impli-
and by society at large. As a result, the German defence cations for their level of ambition and effectiveness.
establishment is either unwilling or dissuaded to pur- It is also an empirical proof of the extent to which a
sue innovation goals in relation to EDTs and auton- strategic culture can inhibit or, in some cases like that
omy which are known or believed to be publicly and of the US, support defence innovation.
politically sensitive and controversial, and innovation This analysis reveals the need to build and pro-
agencies have no independent agency in promoting mote a shared political, military and societal percep-
innovation agendas within the ministry or beyond. 92
tion of national risks and threats and to construct a
However, this is not to say that Germany has not national consensus on these aspects to drive innova-
exhibited innovation in defence closely linked to its tion. In all the five national perspectives this involves
perceptions of its role in European and global security. industry and innovation communities’ activism in
Berlin leads the drive within the EU and NATO for a promoting their solutions,95 intelligence and defence
new and innovative approach to responsible use of EDTs establishments, and national parliaments or execu-
and arms exports. However, the EU-level German-led tive branches communicating effectively.
strategic process on the ‘Responsible Use of Military A societal debate on defence innovation is impor-
Artificial Intelligence’ is not making progress against tant to build a common understanding of the goals
deep disagreements among member states, while the
93
of innovation and measures of efficiency, as tech-
2021 NATO AI strategy includes generic principles of nological progress and innovation hardly exist in a
responsible use, and the US-led Partnership on AI does cultural and political vacuum.96 In addition, as dem-
not function as a forum to define a common set of prin- onstrated by current debates about the importance
ciples of responsible use for AI. Nevertheless, Berlin still of strict screening of Chinese foreign direct invest-
hopes this will become a successful example of German ment (FDI), adversarial technological progress is
defence policy innovation and a German brand as leader met not just with a technological response, notably
in responsible and values-driven defence innovation. 94
investments in technology standards, but also with a
There are strong domestic and cultural reasons behind cultural response, such as the protection of science-
this prioritisation, though. Berlin views clear principles and-technology cooperation, trade and non-defence
of responsible use of technological innovation as a pow- telecommunications.
erful segue in fostering wider relations with German pri- There is strong US bipartisan political support for
vate industry and academia, who are reluctant to work defence innovation, as demonstrated by congres-
with the defence establishment. All the same, it is worth sional steering on the matter through the National
underlining that such a focus on the normative (ethical, Defense Authorization Acts. Public opinion also
legal and moral) aspects of defence innovation appears to largely shares the government’s threat perception
be disconnected both from considerations of their impact of China and Russia.97 However, this requires a sus-
on future national operational advantage and from his- tained effort to raise awareness and inform society
torical analyses of how decision-makers previously dealt and policymakers of the multiple challenges posed
with such normative issues around new technologies. by China.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 15
 Figure 2. Comparative assessment of key defence-innovation drivers across five national perspectives
Drivers / Country United Kingdom France Germany United States China
Threat and vulnerability assessment
Coherent threat and vulnerability assessment and prioritisation1 Yes Yes Yes Yes Yes
Clear innovation priorities and linked to strategic Yes Yes No Yes Yes
competitiveness, warfighting advantage2
Alliance commitments Yes Yes Yes Yes No
Ongoing major military operations No Yes No No No
Political-military-societal support
Convergent political-military-societal support for defence Partial Partial No Partial Yes
innovation3
Political steering of defence-innovation activities and goals4 Yes Yes No Yes Yes
Innovation organisation and governance
Defence-innovation strategy and organisational development5 Yes Yes No Yes Yes
Innovation embedded in main defence establishment6 Partial Partial No Yes No
Organisational and educational innovation culture, openness to Partial Partial No No Unknown
new ideas
Adapted human-talent acquisition, retention and upskill No No No Partial Unclear
programmes
Institutional self-assessment and lessons learned7 No No No No Unknown
Lab-to-battlefield innovation implementation plans/pathways Unknown Yes No Yes Unknown
Active systems of managing failed research Unknown Unknown Unknown Yes Unknown
Broadening defence engagement with private industry and Yes Yes Unclear Yes Yes
academia
Innovation investment
Dedicated defence budget for EDTs Yes Yes No Yes Unknown
R&D&I investment trends in defence 8
Increasing Stagnating Unknown Increasing Unknown
Agile procurement, acquisition practices and DevSecOps No No No No Unknown
Tolerance of failed innovation 9
No No No Yes Unknown
‘Fail small, fail fast’ culture10 No No No No Unknown
Mix of incremental and disruptive innovation 11
Partial Partial Unknown Yes Unknown
Access to foreign research and technology Yes Yes Yes Yes Yes
Strict technology export regulation Yes Yes Yes Yes Partial
Screening mechanisms for FDI in critical technologies Partial Partial No Yes Yes
Source: Compiled by the authors based on the data collected for and presented in this paper

Key

1. Indicated by threat and vulnerability identifica- assessment that is also reasonably shared by the defence establishment structures and processes,
tion in public and defence strategic documents, political class and society more broadly. There is and their subordination to high-level decision-
including national-security strategies, national no major political or societal opposition/refutal of making authorities who have the ability to rapidly
defence strategies, integrated reviews of security identified threats and/or vulnerabilities or of the adopt and scale innovation solutions.
and defence, etc., as well as by mature intelligence role of the armed forces in engaging broadly with
and foresight capabilities to enable early threat society for defence-innovation purposes. 7. Refers to the existence and active use of neutral
identification which increases the need for inno- processes for institutional self-assessment and
vation solutions. Political and military authorities 4. The political leadership is actively involved in incorporation of lessons learned from innovation.
share a clear identification of the (main) adversary. defining, shaping and promoting defence-innovation
goals, processes and structures and in promoting the 8. Based on the evolution of R&D/R&T (research
The wider and more diffuse the range of threats and technology) spending within the context of
identified, the less clear adversary identification implementation of innovation solutions.
national defence budgets over the past 5 years.
and prioritisation is likely to be. 5. Refers to the development and implementation
of strategic documents, including national defence 9. Based on known information about the cancella-
2. A set of agreed high-level innovation priorities are tion of R&D programmes of record.
clearly identifiable based on the identified threats strategies, defence-innovation strategies, new oper-
and vulnerabilities. These are linked to the goal of ational concepts and doctrine, changes in military 10. Based on public information about the cancel-
creating a strategic and/or operational advantage in structure, dedicated defence-industrial strategies lation of innovation projects, proofs of concept and
the competition for global power as well as on the and specific defence EDTs strategies as well as the prototypes after unsatisfactory performance.
battlefield. creation of dedicated institutions, structures, posi-
tions and practices related to defence innovation. 11. Based on the disaggregation of R&D&I invest-
3. Among political and military leaderships there ment by type of innovation (incremental vs disrup-
is a shared sense of need for defence innovation, 6. Refers to the wide integration of defence- tive) and type of partners (traditional defence
based on a convergent threat and vulnerability innovation structures and processes into main primers vs new industry, start-ups and academia).

16 The International Institute for Strategic Studies

 European middle powers remain hard pressed between convergent political, military and societal support can only
a shift towards defence innovation and the need to fill per- be realised if defence innovation is organised and managed
sistent capability gaps. While the number and scope of adequately. As Figure 3 shows, China, France the UK and
European military projects on EDTs is on the rise, many are the US have well-organised defence-innovation systems
linked to incremental innovation plans that seek to either with mature or maturing governance. This includes EDTs
fill national capability gaps or improve the readiness and and defence-industrial strategies,100 new governmental
availability of military capabilities for ongoing operations.98 innovation agencies,101 laboratories and technology accel-
In the UK and France, the military and political leadership erators, dedicated defence strategies for engagement with
are relatively in sync when it comes to national defence- start-ups and small and medium-size enterprises, dedi-
innovation priorities. Beyond the need to hedge for the loss cated innovation funds, adapted organisation structures to
of combat mass, defence innovation also fits in the post- include chief information and data scientist positions, and
Brexit political narrative of an innovative nation. This is not ethical governance. Germany has embraced this challenge
entirely about technological innovation, as the UK’s self- with some delay but has now also created at least the foun-
definition as a science-and-technology superpower is not dations for innovation management.
firmly linked to its rating in S&T or material and industrial The different character of the Chinese political regime
capacity rankings. Rather, it is linked to status aspirations ensures a more coordinated top-down political–military
based on compensating gaps in S&T and industrial capac- consensus, particularly on threat assessment and inno-
ity with policy innovation, such as fostering an ‘honest bro- vation priorities linked to operational advantage. By
ker’ approach to cooperation in new frontier domains and virtue of this, the Chinese political leadership plays a
technologies, including space. France, meanwhile, is more disproportionate role in steering defence-innovation
invested in developing its industrial and S&T capacity as a plans at the expense of bottom-up innovation efforts.
way of ensuring its ‘strategic autonomy’. However, ongo- Xi Jinping emphasised the importance of disruptive
ing military operations in the Sahel skew France’s priorities technologies for a new revolution in military affairs
and may still affect the focus and investment on defence and has charted a course for the Chinese armed forces
innovation given tight material resources. towards becoming a ‘world leading military’ by 2049.102
Moreover, political uncertainty in view of recent and The centralised nature of the Chinese political system
upcoming elections in Germany and France still has the means Beijing is better placed to exploit a strategy of
potential to undermine political support for innovation military–civil fusion. The latter is central to China’s
plans. Germany is perhaps the best example of how a ability to derive military advantages from domestic
lack of convergent political, military and societal sup- and foreign commercial technological and science rela-
port inhibits defence innovation. It also deprives innova- tions achieved through both legal and illegal means.103
tion efforts of clear political steer, connected to long-term Western analysts fear China’s military–civil fusion
national strategic and operational interests. This contrib- may amount to a strategic advantage in technological
utes to a feeling that Germany is ‘sleep walking through competition.104 The US National Security Commission
the innovation process’.99 As demonstrated by the recent on Artificial Intelligence assessed China’s and America’s
general-election campaign, foreign and defence policy is competitiveness in AI along six technology stacks – data,
not a primary concern for German policymakers, or for human talent, algorithms, applications, hardware and
society more broadly. And the deeply rooted German stra- integration – and concluded that the two countries were
tegic culture rejects a more prominent role for the military quite evenly matched, but that in the next few years one
in sensitive technologies such as AI or in the development of them may emerge with a technological advantage. The
and use of autonomy in weapon functions. commission’s assessment is that while China leads in
applications, integration and data, given its large popula-
3. Innovation organisation and governance tion and unincumbered legal environment to harvest and
The added value derived from the prioritisation of innova- use data, the US leads in human talent, algorithms and
tion goals, a shared threat and vulnerability assessment and hardware. Though such extreme opinions may not be

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 17
representative of the US administration’s views, former and security considerations. This is evident in the limita-
Pentagon officials claim the US has already lost the AI tions on the exploitation of private data, though the US is
race with China. 105
Democratic regimes require a much actively investing in technological solutions for low-data
more delicate balance between rights and civil liberties algorithms.

Figure 3. Comparative timeline of key strategy and institutional developments in defence innovation across five national
perspectives (2015–present)


March 2015 October 2017
UK Quantum National Strategy China decides to build ‘Quantum Research Supercentre’
backed up by US$10bn investment
May 2015
December 2017
China’s Military Strategy
Made in China 2025 Strategy US National Security Strategy
China’s Three-year Action Plan for Promoting Development of a
New Generation Artificial Intelligence Industry (2018–20)
July 2016
Germany’s Defence White Paper on Security Policy and the January 2018
Future of the Bundeswehr
US National Defense Strategy

September 2016 France’s Military Cyber Strategy

UK launches the Defence Innovation Initiative and establishes

March 2018
Defence and Security Accelerator (DASA)
France’s AI Strategy
October 2016
April 2018
US Preparing for the Future of Artificial Intelligence
US National Artificial Intelligence Research and Development UK AI Sector Deal


Strategic Plan
May 2018
November 2016 US White House Summit on AI
Germany’s Cyber Security Strategy AI in the Nordic-Baltic region
Germany’s Hightech Strategy 2025
December 2016
US Artificial Intelligence, Automation, and the Economy report August 2018
UK Data Ethics Framework
March 2017
Pan-Canadian AI Strategy Sepember 2018
US Army Robotics and Autonomous Systems Strategy France establishes l’Agence de l’innovation de défense (AID)
France’s Defence Innovation Guidance Document
April 2017
Source: IISS
Germany’s Guideline on Digitisation November 2018
Germany’s AI Strategy
July 2017
Bundeswehr creates the Cyber Innovation Hub December 2018
China’s New Generation AI Development Plan US National Military Strategy
US launches National Quantum Initiative


Key
Canada/US France/Germany /UK China/Russia

18 The International Institute for Strategic Studies

 February 2019 February 2020


US White House Executive Order No. 13859 on Maintaining US adopts ethical principles for AI
American Leadership in Artificial Intelligence Germany’s Strategy Paper of the Federal Government on
US Department of Defense Artificial Intelligence Strategy: Strengthening the Security and Defence Industry
Harnessing AI to Advance Our Security and Prosperity
UK Science and Technology Strategy

May 2019
September 2020
US Senate bill proposal on Artificial Intelligence Initiative Act
DoD/JAIC launches the AI Partnership for Defense
Beijing AI Principles DoD/JAIC launches early version of Joint Common Foundation
(JCF)
June 2019
October 2020
Russia’s national AI strategy
US National Strategy for Emerging and Critical Technologies
July 2019 US National Quantum Strategy
UK Science and Technology Strategy
US National Security Commission on AI Initial Report to
DoD Data Strategy
Congress
DoD Electromagnetic Defense Strategy

August 2019
February 2021
US NIST ‘U.S. Leadership in AI: A Plan for Federal Engagement
in Developing Technical Standards and Related Tools’ DoD/JAIC AI Ethical Principles
UK Intelligence Agency publishes AI and Data Ethics
September 2019 Framework
Germany’s strategy paper on the Future of the Security and
France’s Artificial Intelligence in Support of Defence report Defence Industry (including EDTs)
UK Defence Innovation Priorities Paper Germany’s Position Paper: Reflections on the Bundeswehr of
UK Defence Technology Framework the Future


October 2019 March 2021

US Defense Innovation Board guidelines for safe and ethical AI UK Defence in a Competitive Age paper
Recommendations of the Data Ethics Commission for UK Defence and Security Industrial Strategy
the Federal German Government’s Strategy on Artificial
Intelligence China’s PRC adopts 14th Five Year Plan

November 2019 April 2021

US Army Annex to DoD AI Strategy France’s Strategic Update 2021

Germany’s position paper AI in the German Land Forces UK Integrated Review of Security, Defence, Development and
Foreign Policy
US National Security Commission on AI Interim Report to
Congress on AI
May 2021
Russia announces US$790m investment in quantum technology
over the next five years Germany’s Key Elements of the Bundeswehr of the Future
German government invests ~ US$1bn in quantum technology,
December 2019 including with the aim of building a quantum computer

US Air Force Annex to DoD AI Strategy

June 2021
Chinese decree orders state offices to replace foreign PCs and
software US Joint Chiefs of Staff approve new Joint All-Domain
Command and Control (JADC2) Concept
January 2020 Germany pledges to invest another US$2bn in quantum tech
by 2025
UK Strategic Intent on Quantum Technology
French Parliamentary Report on Quantum Technology and September 2021
Strategy
France launches AID-AIT partnership for U space programme
France launches the Innovation Défense Lab (within AID/DGA)


Source: Compiled by the authors from multiple sources

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 19
Embedded and structured defence-innovation prioritising defence innovation within the Department
systems of Defense stems from the patronship of the Office of
Rather than focus on purely quantitative metrics of the the Secretary of Defense over several of the critical
organisation of defence innovation, empirical data sug- innovation platforms, including JAIC, DIU, SCO and
gests qualitative metrics offer enhanced explanatory DARPA, combined with higher flexibility of implemen-
power, particularly of the efficiency of national defence- tation. These institutions fulfil very different roles in the
innovation systems. Two such qualitative metrics are national defence-innovation system.108 However, they
consequential for the efficiency of innovation systems. share several common features: a direct link with high-
The first is embeddedness and adaptation to national level decision-makers (either the secretary of defense or
and organisational cultures, strategic contexts and inno- the relevant under-secretaries of defense) who have the
vation goals. Far from being exclusively an asset, China’s legal authority to transfer innovative solutions to the
centralised and politicised system, consisting of multi- military users and scale them across the defence estab-
ple hierarchical layers of bureaucracy, may act to slow lishment, various degrees of organisational flexibility
down Chinese innovation. Among its key weaknesses related to procurement and contracting in relation to
are inaccurate reporting, lack of mature experience and non-defence traditional partners, and embeddedness
expertise in integrating frontier technologies, and an within the defence enterprise, demonstrated by the
overwhelming top-down direction of innovation which greater focus on emerging DevSecOps models, rather
may not fully integrate the needs of the military users than representing a separate track.
and operators. 106
The benefits of centralised planning are Qualitative differences in how adapted the innova-
potentially offset by the lack of bureaucratic agility and tion organisational structure is to political and organi-
are highly dependent on the quality of the political lead- sational cultures are critical, emphasising the limits of
ership that steers defence-innovation efforts. emulation. For example, there is a fundamental differ-
However, this is not exclusively a Chinese challenge. ence between the US and European powers in gener-
The American innovation system is perhaps the best ating alternative sources of funding for software- or
developed out of all the five nations. It comprises a wide- hardware-driven defence innovation.109 While the US
ranging variety of research and innovation facilities, and the UK are more challenged in identifying venture-
with varying degrees of centralisation and embedded- capital funding for hardware-based and manufacturing
ness within the Department of Defense ecosystem. Such innovation, France and Germany are more challenged
facilities include the National Research Foundation, in identifying venture-capital funding for software-
Defense Advanced Research Projects Agency (DARPA), based innovation. This proves that defence-innovation
military research labs, dedicated centres like the Joint best practices and procedures are not immediately
Artificial Intelligence Centre (JAIC), the Defense transplantable even in the case of close allies.
Innovation Unit (DIU) and the Strategic Capabilities Nevertheless, the US defence-innovation ecosystem
Office (SCO), and public private partnerships, includ- is a benchmark both for allies like the UK, France and
ing government sponsored university research labs. Yet Germany as well as for strategic rivals like China. Either
their efficiency is not directly related to how centralised bilaterally or through NATO, the US exercises a great
or decentralised they are nor whether they foster top- degree of influence over the organisation of defence
down or bottom-up innovation. 107
innovation by its European partners, for whom emulat-
Rather, empirical evidence suggests the efficiency of ing American innovation structures and best practices
innovation structures is influenced by their embedded- is often appealing. This is demonstrated by the French
ness in the defence ecosystem at the right level of deci- and German calls for the development of DARPA-like
sion-making authority to facilitate adequate strategic innovation structures, as well as their establishment of
steering of innovation efforts, sustainable investment such agencies at the national levels.
and rapid transition of innovation solutions from the Having failed in leading the way in the establishment
lab to the battlefield. For example, part of the success in of an EU-wide DARPA-like innovation agency,110 in 2019

20 The International Institute for Strategic Studies

President Macron launched l’Agence de l’innovation de armaments and cyber directorates – which do not
défense (AID), embedded with the Direction Générale necessarily coordinate and cooperate with each other.
de l’armement (DGA) in the French Ministry of Defence, German experts also point to the remarkable discon-
and in 2021 Paris pushed for enhancing the role of the nect between the innovation and technical experts and
European Defence Agency as the EU-level leading the decision-makers within the Bundeswehr and the
defence-innovation agency. 111
Similarly, the German German Federal Ministry of Defence.113 Consequently,
federal government established the Cyber Innovation the political level in the ministry is reluctant to follow
Agency (ADIC) whose role was initially to accelerate and implement innovation solutions proposed by inno-
and upscale innovation in EDTs, but was subsequently vation teams, and the Bundestag seeks strict budget-
limited to innovation in relation to cyberspace, notably ary oversight of any defence project over €25 million.
to ‘use explicitly risk-taking methodology to initiate Meanwhile, there is a more substantial calibration
ground-breaking and ambitious cybersecurity technol- and coordination problem at play in the German sys-
ogies and other key technologies that can create strate- tem. While the outputs of innovation agencies like
gic advantages for internal and external security and to the BMVg Futures Analysis Branch or the Fraunhofer
further develop them for the needs of the Bundeswehr Institutes and Research Units are usually classified,114
until they are ready for use’.112 Meanwhile, the UK limiting broader engagement with industry and aca-
has developed a dense ecosystem of R&D&I agencies, demia, the Cyber Innovation Agency does not possess
including the Dstl, the UK Strategic Command’s JHUB, any legal recourse to attribute classified projects or de-
the Ministry of Defence’s new AI Centre and the new risk technologies it is working on for the Bundeswehr.
civil-military Technology Accelerator, which cover both In 2019 the Bundeswehr’s research centre published
front and back-end innovation comprehensively, in a detailed position paper on ‘Artificial Intelligence in
similar ways to the US system. Land Forces’, focusing on three objectives: increased
Nevertheless, both in the French and German cases, efficiency in routine duties, improved capabilities on
their innovation agencies lack appropriate direct links operations and addressing potential capability gaps.115
to political decision-making authorities to ensure their However, it is unclear whether the process led to any
innovation solutions are rapidly adopted and scaled. outcome with the Bundeswehr or the Federal Ministry
This also limits the ability of the defence-innovation of Defence.
community to actively promote the implementation By contrast, under the French DGA, AID’s role is that
of their innovation solutions with the political estab- of a ‘catalyst for innovation’116 across the whole min-
lishment and society at large. In France, the innova- istry. The AID’s attributions combine innovation for
tion agency depends, including for its budget, on the operational engagement needs, for ongoing capability
defence-procurement directorate, whereas in Germany, programmes as well as more disruptive and founda-
the ADIC is jointly controlled by the federal ministries tional innovation. The AID’s mission is not dissimilar
of defence and of the interior. to the JAIC, though it is more pressed in its funding
Overcoming the stovepiped legacy of innovation allocation between procurement projects and more dis-
systems is a challenge in the US, the UK, France and ruptive innovation ones.117 The AID’s mission includes
Germany, while there is little specific public knowledge providing guidance on defence-innovation priorities,
about how this aspect is handled in China’s defence- piloting and upscaling defence-technology innovation
innovation efforts. For example, the fragmentation projects, including in relation to private industry and
and duplication of innovation attributions and roles academia, and accelerating defence innovation across
in Germany is a ‘stumbling block to innovation’ and the defence enterprise. The AID is efficient in provid-
generates unhelpful competition between services, ing for the innovation needs of the services by virtue of
branches and agencies. Innovation attributions are its wide embeddedness within the defence enterprise,
fragmented among the three main directorates of the and it is well embedded in a growing French defence-
Federal Ministry of Defence – notably, the planning, innovation ecosystem.118 However, it has only limited

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 21
exposure to the ministry’s political leadership to rap- accelerator towards the market, in the space context and
idly transition its technological solutions to the battle- beyond. In France, the AID and its newly established
field and to experiment with operational concepts and Innovation Défense Lab122 have authority to accelerate
force structure. defence-technological projects linked to future capa-
Meanwhile, fragmentation also reveals some of the bilities, innovation projects that correspond to civilian
key challenges in effectively organising bottom-up and technology and foundational research projects,123 albeit
piecemeal innovation. While the US has had a coher- unclear whether the process of transfer and transition is
ent federal approach to autonomy in defence for a dec- different from traditional approaches. In Germany, the
ade or longer, France, Germany and the UK have only process to transition innovation solutions from the Cyber
begun to consider this function more recently. Unlike Innovation Agency or the Fraunhofer Institutes to the
the American, French and British national defence eth- Ministry of Defence for development remains unclear,
ics and governance of defence innovation, the German and the classification of their defence-innovation efforts
Ministry of Defence is reluctant to assume such a role limits their ability to transfer these solutions to the com-
and prefers hybrid alternatives. For example, the min- mercial market if their applicability for defence is limited.
istry prefers to participate as an equal partner in a The peculiarities of the European states’ defence-
bottom-up ad hoc Airbus-led multistakeholder initia- innovation systems create challenges with rapidly
tive that explores ethical challenges and principles of upscaling defence-innovation solutions across the
responsible use of AI technologies and autonomy in defence enterprise. This is in part related to the type of
the FCAS 119
multinational projects, which has limited innovation that is desired for upscaling (notably, disrup-
potential for upscaling. tive innovation is more challenging to upscale rapidly
and encounters more resistance than incremental inno-
Transition and transfer planning for innovation vation), the reliability and robustness of the technologi-
solutions cal innovation solutions and the availability of channels
The second critical metric for the efficiency of defence- to transition innovation solutions to the war fighters.
innovation systems is advanced planning for the transi- Importantly, it is also about the scope of the innovation
tion and transfer of innovation solutions to the military solution itself. Technological innovation related to a capa-
users or the market. In the US, DARPA has long had bility programme will be easier to scale than narrow tech-
technology-to-market and technology-transfer plans nological solutions. Therefore, to increase both impact
towards military users 120
and the Department of and efficiency of transfer, defence innovation is better
Defense is developing fast-track pipelines for its major suited for overarching challenges to national security, be
technology initiatives. Moreover, it has implementation it at the strategic, operational or tactical level, rather than
promotion plans for its innovation solutions and is con- narrow and incremental technological solutions.
stantly engaged with government and industry to push The military–civil fusion in China is generally con-
the implementation of its outputs. 121
This is a rare occur- sidered to be effective in facilitating the transition
rence among national innovation ecosystems, where and transfer of innovation solutions and technologies
innovation agencies usually lack both the remit and the between the defence and civilian domains. However, lit-
planning authority to actively promote the implementa- tle is publicly known about the exact processes through
tion of their solutions. which this transition takes place.124 Nevertheless, the
However, the lack of clear transition pathways, pipe- lack of public information is neither indicative of the
lines and partners from innovation agencies towards absence of Chinese challenges in implementing mili-
military end users (or indeed private industry) is a consid- tary–civil fusion nor a measure of efficiency in transi-
erable challenge for European countries. In the UK, there tioning technology solutions between the military and
are recent indications that strategic thinking is consider- civilian sectors. The Chinese system is faster at direct-
ing more flexible plans to transition some of the research ing and mobilising innovation resources than those in
developed by the Dstl or the civil–military technology Western countries, albeit the military–civil fusion also

22 The International Institute for Strategic Studies

deliberately creates a grey area for the Chinese govern- non-traditional private industry and academic partners
ment to exploit by simultaneously reducing its commer- in pursuit of defence innovation. While access to fron-
cial and defence-industrial disadvantages. 125
However, tier technologies130 and speed of adoption are important
how China manages to transition technology innova- considerations, agile procurement is about a changing
tion in the framework of military–civil fusion between relationship between the defence establishment and
the defence and civilian sectors, how it manages to industrial and academic communities. It is also about
break down barriers between the military and the shifting mindsets on the inherent value of the procure-
defence industry relative to high-tech businesses, and ment process, which is to integrate innovative solutions
how it manages to adapt its procurement practices are reliably and robustly and not just to save funds. Finally,
key challenges for the effectiveness of Chinese defence it is about possessing adapted technological horizon-
innovation, without a predetermined outcome. 126
Far scanning tools that contribute to a better understanding
from needing just a procedural adaptation, the chal- of how new technologies will shape future operating
lenge for China, like for the US and European states, environments and guide governmental decisions about
is to develop and foster a culture of innovation at the when and where to initiate changes in R&D and capa-
centre of an essentially conservative bureaucracy. bility development.
Furthermore, China uses a wide variety of legal and The focus in existing literature is rightly on widening
illegal means to ensure transfer of foreign technologies engagement with industry, particularly non-traditional
and leverage foreign innovation and technology into industry, and academic partners. However, particularly
its own defence-innovation efforts. 127
These include, from a capabilities point of view, this view needs to be
among others, Chinese technology-for-market trade- balanced against the reality that defence primers will
offs, requirements to create joint ventures with Chinese most likely continue to play key roles in systems inte-
companies in return for market access for foreign com- gration. This requires that the defence industry makes
panies, and mergers and acquisitions. The 2016 Kuka more concerted efforts to adapt its supply chains in
acquisition and the recent Chinese investments in the timely fashion, and that government innovation agen-
British semiconductor sector 128
are pertinent exam- cies ensure they are well positioned to facilitate the
ples of such Chinese practices, which put a premium effective networking between these non-governmental
on the development and coherent implementation of relevant stakeholders, while ensuring a viable balance
FDI-monitoring tools in the US and Europe. The lat- between over-fragmentation and over-concentration of
ter is particularly important because a shift in Chinese national defence industries. This issue is under consid-
foreign-investment practices developed over the past eration in the mix of government-owned architectures
decade 129
is unlikely in the foreseeable future. and platforms and privately owned relevant capacity
China’s quest for access to foreign technology that – such as in military cloud and edge computing, space-
compensates for its own shortcomings is of a different domain awareness and others. Moreover, new tech
nature and scope. Nevertheless, as demonstrated by actors on the defence market offer innovation-as-a-service
recent debates in Washington and across Europe, the packages. This entails higher risks in developing new
US and European middle powers also struggle with products which may be relevant for and commercial-
identifying the right balance between access to foreign ised to defence institutions across the world. However,
technology and imposing strict export regulations that they are not the end-result of a development process
limit the diffusion of innovation and the potential com- driven by government specifications but rather by the
petitiveness gains that come with it. tech company’s own marketable product. One such
example is Anduril Industries, which sells subscrip-
Agile procurement tions to its Lattice AI software to several countries. In
This lack of clear technology and innovation transition 2021, the UK signed a new type of contract with Anduril
pipelines has implications for procurement and acquisi- Industries whereby it subscribes to the Lattice software
tion processes and the ability to engage broadly with to enhance its army’s command and control and force

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 23
protection.131 In 2020, the Munich-based Bundeswehr administrative environment. Its Defence Innovation
University organised a Defence Innovation Pitch Day Fund, which already suffers from limited financial
which prominently featured technology start-ups, while resources, is legally required to invest in civilian technol-
Behörden Spiegel has an online ‘Defence Innovation Talk’ ogies that have already proven some level of success on
series in which companies like Anduril Industries are the market.137 However, this may be too late. In response,
promoting their innovation-as-a-service model.132 AID is piloting a ‘contracting factory’ which explores dif-
Empirical evidence suggests an increasing flexibility of ferent acquisition and contracting methods that are more
procurement and acquisition in national practices, within appealing to non-traditional industry partners to facili-
the realm of procurement law and beyond. In the US, tate rapid access and uptake of new technologies. While
the JAIC and DIU are now widely using so-called ‘other initially developed for the FCAS project, France’s digital
transactions authority (OTA)’, a limited and previously design, manufacturing and services (DDMS) uses smaller
exceptional budgetary authority for the Department and modular procurement packages to speed product
of Defense to pursue specific prototypes, research and development at Airbus and Dassault138 and has the possi-
production projects that are linked to national security bility of being upscaled. By contrast, the implementation
and technological superiority goals. 133
Software- rather of German procurement law is very strict in process and
than hardware-based innovation solutions also require conditionality on engagement with (defence) industry to
different acquisition and development processes. The avoid assumptions of undue influence.139
US Defense Innovation Board (DIB) recommended as
early as 2017 the expanded use of acquisition waivers Managing failed innovation
and exceptions 134
as a way of ensuring development and Defence innovation inevitably involves a degree of fail-
acquisition practices for software products are agile and ure. The relative lack of transparency and classification
iterative. In cyber and space domains, where non-state of R&D projects in the innovation sphere is an important
actors play a growing role and have greater agency, there challenge in assessing the efficiency of national innova-
is more pressure to revise traditional customer-supplier tion efforts. It is also an important stumbling block for
relations with industry towards more balanced partner- innovation and a symptom of nations’ reluctance and
ships that account for the (security) interests and needs unease in managing failed innovation – and its public
of all stakeholders. accountability implications. The quantitative metrics
European states have less flexibility in their pro- of failed innovation are sometimes staggering. While
curement and acquisition processes, especially when decisions to cancel programmes of record are always
innovation is still predominantly linked to long-term politically challenging, an analysis of the top ten failed
defence-modernisation cycles. 135
In the UK and France, American R&D projects reveals costs of over US$53bn
the Defence and Security Accelerator (DASA), the newly overall in two decades and individual project costs of
established Defence AI centre, the civil–military technol- more than US$2bn.140 The flexibility within DARPA to
ogy accelerator, and the AID have higher flexibility in fund frontier research and periodically assess its pro-
contracting powers and can rapidly contract projects jects through projections on chances of implementation
from within the defence establishment and from out- means DARPA already has a mature culture of failing
side it. The UK is developing a centrally controlled small and fast.141
experimentation enterprise for defence, including using There are no similar publicly available metrics for
synthetic environments, 136
which could become an European countries, where the transparency of defence
effective and accelerated transition pathway for techno- R&D expenditure remains limited. Moreover, while US
logical and operational innovation solutions to the bat- defence spending on R&D has reached historical highs
tlefield. The AID in France faces challenges to rapidly since 2018, increases in European defence spending did
identify the financial means to transition a technologi- not translate into proportional increases in R&D, which
cal innovation towards a final product and end users in still lags behind 2007 levels.142 This makes it challenging
the Ministry of Defence because of a dense and complex to understand how European countries manage failed

24 The International Institute for Strategic Studies

innovation. However, a bird’s eye view of European Implementing a flexible oversight framework which
R&D spending suggests known R&D programmes gen- intensifies as technological solutions mature, as their
erally have longer lifespans despite cost and schedule costs grow or as they transition to capability develop-
overruns, and the fact that crossing the ‘valley of death’ ment is one such business practice. Incorporating early
is a far greater challenge than in the US. In other words, prototyping into innovation projects before the bench-
Europeans neither fail fast nor small, and therefore marks for continuation is a staple of technology innova-
require a significant shift in their approach to innova- tion processes in the main tech-industry actors,146 and
tion failure so as to be able to terminate unsuccessful a proven instrument of more efficient management of
projects early and flexibly. innovation and limited resources for higher returns. This
Managing failed innovation is not just about higher tol- is an important de-risking strategy for more disruptive
erance thresholds to innovation failure risks. Importantly, technologies which has proven useful in the private sec-
it also requires ensuring recurring institutional self- tor. However, this also requires a re-evaluation of test-
assessment and lessons-learned processes in relation to ing, validation and certification practices to keep up
innovation outputs, particularly in relation to the mix with technological developments. Innovation agencies
of incremental and disruptive innovation investments, need more testing and evaluation capacity than is avail-
which we will discuss in the last section of this paper. able today. While digital engineering and digital twins
The more disruptive and foundational the technology, for technological development are already mature and
the higher the likelihood of innovation failure. Yet estab- widely used practices in the private and defence indus-
lishing processes of self-assessment and internalising les- try, government certification and validation processes for
sons learned from failed innovation is critical. technological innovation still depend on physical proto-
By way of an example, DARPA is often portrayed typing, which is both lengthier and more expensive. The
as ‘a proven model in the innovation space’, with an UK has announced an intention to scale the use of syn-
established track record of successful innovation. 143
thetic environments for experimentation, training and
However, the examination and assessment of its success testing to circumvent some of these challenges, whereas
metrics remains challenging. DARPA’s Assault Breaker the US regards this solution as a possible answer to mili-
project reportedly failed to achieve the initial stated tary reluctance to integrate new technologies.147
goals of developing a fully networked precision strike
capability.144 Yet it was instrumental in the development Human-talent pipeline
and implementation of Joint Surveillance Target Attack Any measure of efficiency of defence innovation largely
Radar System (JSTARS), which linked a variety of pre- depends on the availability and quality of human talent
cision-guided munitions and platforms with advanced in the innovation infrastructure, the armed forces and
Intelligence, Surveillance, and Reconnaissance capabili- the ministries of defence. However, human talent with
ties 145
and paved the way for network-centric warfare. STEM and technological expertise is in short supply
This example highlights the need to critically assess across all the five countries analysed here. A measure
(seemingly) failed innovation and internalise the les- of the challenge is, for example, the difference between
sons learned from such failures to understand future the EU’s investment to develop 20 million EDTs spe-
innovation investments. Moreover, as dual-use techno- cialists by 2030 and its estimate that, if all works well,
logical development goes both ways between the gov- it may have as many as 12 million – leaving it 8 mil-
ernmental and the commercial sector, a reconsideration lion (or 40%) short of estimated need.148 This situation
of the metrics of success for innovation, including in is comparable in scale across all the five national per-
defence, is warranted. spectives presented in this paper. In addition, except for
Business practices in technological innovation are a China, all four other countries face significant negative
potentially useful example for government-led efforts. demographic trends that limit the pool of suitable new
Developing a mix of incremental and disruptive innova- recruits. Furthermore, the adaptation of talent acquisi-
tion in R&D&I portfolios is key to tech-industry success. tion, retain and upskill processes in government has

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 25
been slow. Innovation does not feature in military edu- American R&D funding remained steady after the end
cation programmes in any of the five countries stud- of the Cold War153 before beginning to rise in the last
ied, which makes the challenges of fostering a culture three years and is expected to reach an all-time high in
of innovation – a priority in the US and in the UK, for 2022.154 The Chinese defence budget has grown con-
example149 – significantly more challenging. stantly for three decades,155 though very little is known
Upskilling the workforce in defence establishments about how the Chinese government and military priori-
is essential to effective defence innovation. This entails tises between, for example, filling persistent mechanisa-
the review of governmental procedures to attract and tion gaps and defence innovation. However, the peace
retain new talent, including a re-evaluation of inflexible dividend marked a cliff rather than a dip in defence
and lengthy security-clearance procedures, the mini- R&D spending in France, Germany and the UK – one
mum qualification levels of new recruits and considera- from which these states were yet to fully recover even
tions about the contribution that highly skilled civilians before the COVID-19 pandemic reintroduced concerns
who are not apt for military service may nonetheless about the sustainability of European defence spending.
make to the defence enterprise. It also entails a more Latest defence-expenditure data indicates that R&D
applied shift in defence establishments towards skill- and particularly R&T spending in Europe is still below
biased technological change and understanding where 2007 levels.156
this new expertise can be leveraged within the institu- France pledged to invest annually approximately
tional chain. In other words, highly qualified human €100m in defence-related AI research between 2019 and
resources in the armed forces are an enabler of more 2025.157 The AID’s budget is nearly €1bn, though a por-
rapid innovation and integration of new technologies. tion of it is appropriated to existing procurement proj-
For example, recent data-driven research suggests the ects rather than innovation initiatives.158 By contrast to
1990s revolution in military affairs led to a pattern shift British innovation agencies like the Dstl, AID is not an
in US military recruitment towards more highly skilled independent legal entity and its funding and contract-
volunteers, 150
which in turn facilitated greater and more ing authority derives entirely from the DGA, with little
rapid digitalisation. Investment in military education flexibility under French procurement law. Moreover,
serves to further complement this asset. according to the French Cour de Comptes, AID’s opera-
Access to foreign talent is as important as access tional costs are disproportionally high by comparison
to foreign technology, and flexible immigration poli- to its investments, although that report was issued
cies tend to favour such wide access to foreign human only two years after the launch of AID, when the entity
resources in national defence-innovation efforts. While was still developing.159 In a recent report, the French
France, Germany, the UK and the US vary in this Assemblée Nationale criticised the lack of transparency
respect, China is deploying innovative channels of around AID’s funding and contracting instruments as
attracting and retaining foreign talent, albeit selectively. an obstacle to innovation, measuring their efficiency and
An example is the Chinese practice of acquisition-hire circumventing parliamentary oversight.160 Furthermore,
to secure foreign STEM and engineering talent, as in the AID and the French Ministry of Defence are hard
the case of Alibaba’s acquisition of the German start- pressed by difficult trade-offs between sustaining readi-
up Data Artisans, 151
in addition to generous grants and ness and the high operational tempo in the Sahel or
fellowship positions with Chinese-based research enti- investing in longer-term innovation. In Germany, the
ties. 152
Nevertheless, as China’s demand for STEM talent Cyber Innovation Agency launched in 2020 had an ini-
grows, so will the pressure on its immigration policy to tial budget of €350m, though there are doubts about the
attract relevant talent in sufficient numbers. short- and long-term sustainability and overall suffi-
ciency of this budget.161
4. Defence-innovation investment In November 2020, the UK announced a 34.8% increase
Finally, the fourth key driver of defence innovation is in the defence R&D budget (a £1.5bn year-on-year
sustainable and sufficient funding over the long term. increase)162 and the establishment of a dedicated budget

26 The International Institute for Strategic Studies

for emerging technologies. Dedicated national fund- partners. Nevertheless, as procurement practices change
ing instruments have also been created, from national and broader engagement with non-traditional part-
venture-capital funds to defence-innovation funds. ners emerges across these five countries, it is important
One such example is the £66m Transformation Fund to ensure an appropriate level of funding innovation
launched in 2019 to fast-track military robotic projects through partnerships with a wide range of stakeholders.
onto the battlefield. 163
Nevertheless, creating the financial For example, DARPA spends nearly 40% of its budgets
envelope for these new investments in defence innova- on non-defence-industry partnerships.168 In the UK the
tion requires a painful trade-off of higher vulnerability Dstl spends 44% of its budget on research,169 including
from lower numbers of conventional military platforms high-risk research. By contrast, in France, 94% of AID-
available before the entry into service of next-generation funded innovation projects are in partnership with pri-
capabilities generated by defence-innovation efforts. vate industry,170 including defence primers.
These challenges raise doubts about the scale of inno- Furthermore, empirical evidence from national
vation impact and the disruptive nature of its outputs. defence-innovation systems like the US and from the
This is particularly concerning because France, Germany private sector suggests that innovation systems thrive
and the UK represent over 85% of defence R&D spending on a mix of incremental innovation linked to capability
in Europe and over 50% of European investment in AI development and disruptive innovation linked to cre-
and other EDTs. France, Germany and the UK attract 80% ating strategic advantage. Incremental defence related
of AI start-up funding in Europe and nearly two-thirds of to legacy platforms or next-generation capability pro-
the defence-industry R&D investment in Europe, 164
albeit grammes under development will continue, but a pri-
most of this private funding is of national origin. 165
oritisation of defence innovation requires a balanced
The US faces similar challenges and trade-offs. The approach between it and more foundational and dis-
Third Offset Strategy was launched during sequestra- ruptive innovation aimed at ensuring sustained opera-
tion, when US military commanders and political lead- tional advantage.
ers were focused on the recapitalisation of US military Business practices again may be a relevant example.
power and readiness levels in relation to major ongoing Big tech companies deploy an 80%/20% ratio between
military operations in the Middle East. This detracted incremental and disruptive innovation,171 a distribution
for years from the prioritisation of defence-innovation similar to that of the US Department of Defense’s R&D
efforts. Even nowadays the precarity of congressional innovation initiatives, albeit there is no inherent plan-
budgetary politics creates long-term uncertainty ning or directed investment mix between incremental
around the Department of Defense budgetary commit- and disruptive innovation within the Department of
ment to defence innovation. While the Trump adminis- Defense or the US military services. In Europe most
tration was wedded to preserving a 355-ship US Navy defence R&D is linked to incremental innovation and
(prioritising the number of platforms), the implementa- ongoing capability development projects172 – this has
tion of the Biden administration’s shift towards inno- also been the case in the US until very recently. For
vation remains uncertain. 166
Furthermore, democratic example, of AID’s 576 projects, 87% are accelerated inte-
limits and restrictions in data collection for military use gration of innovation projects which are linked to pro-
entail risks that the US may lose its relative technologi- grammes of record rather than disruptive-innovation
cal advantages 167
in critical technologies like AI. projects.173 Even more importantly, developing greater
Beyond the quantitative metrics of investment in flexibility to allocate funds from incremental towards
defence innovation, it is worth considering some quali- disruptive innovation will be increasingly critical mov-
tative aspects, too. For example, the diversification ing forward. The US is currently developing processes
and disaggregation of investment-spending patterns is to enable greater flexibility in their defence-innovation
important. Defence-innovation agencies traditionally funding, but none of the European powers seem to be
channel most of their funding towards defence-industry working in this direction.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 27
Concluding Remarks

This paper finds that a well-structured national sys- transition-and-transfer pathways clear towards
tem for the long-term management and organisa- both the defence enterprise and the market.
tion of defence innovation is a critical element of its  Focus defence innovation on overarching
efficiency. A balanced quantitative and qualitative challenges to national security, be they at the
assessment is needed for a more accurate understand- strategic, operational or tactical level. Defence-
ing of innovation efforts and their efficiency in differ- innovation efforts informed by clear threat
ent national contexts. and vulnerability assessments and seeking a
The paper identifies four key drivers of defence degree of ‘creative disruption’ prioritise high-
innovation across China, France, Germany, the UK level and strategic challenges rather than nar-
and the US: threat and vulnerability perceptions; con- row and incremental technological solutions.
vergent political, military and societal support for  Actively advocate for and promote the imple-
innovation; innovation organisation and governance; mentation of innovation solutions. Facilitate a
and investment in innovation. We find the following role for defence-innovation agencies to actively
criteria to shape and influence the national defence- pursue this for their innovation solutions.
innovation efforts across these countries:  Foster political and societal support for defence-
innovation efforts. Cultural, political and soci-
 Be highly adapted to national and organisa- etal barriers to innovation are just as important
tional contexts, cultures, and vulnerability as intra-military and technological ones.
and threat perceptions. Emulating best prac-  Create dedicated strategies for managing
tices, institutions and projects from other failed innovation and incorporating lessons
international actors while adapting them learned. Most disruptive innovation is neither
to national contexts may be useful for both a very rapid nor a stand-alone process and
great and middle powers. As this analysis has it requires cumulative knowledge and tech-
proven, China, France, Germany and the UK nology developed over time, in one or mul-
all regard the US as a benchmark in terms of tiple projects, some of which may partially
defence innovation. or entirely fail. Nevertheless, an assessment
 Be embedded in the main national defence mechanism for failed innovation can ensure
structures and processes and be linked to valuable lessons, technologies and ideas are
high-level decision-making authorities in not wasted.
the ministries and services that can cham-  Establish dedicated processes to attract, retain
pion and rapidly transition new technologies and train a mix of human talent and adapt
and innovation solutions to adoption and scal- military education to include innovation.
ing. Develop anticipated planning and make  Ensure sustainable long-term investment in

28 The International Institute for Strategic Studies

defence innovation, including by encouraging While it is premature to determine whether the
halo effects
174
wherever possible and pursuing national approaches analysed here amount to models of
a balanced mix of incremental and disruptive defence innovation, or to pass judgement on their level of
innovation projects in defence planning. This success, this paper has sought to analyse and conceptu-
mix should include investments in incremental alise the key drivers of defence innovation and how they
innovation linked to ongoing capability projects are linked to greater efficiency and operational advan-
as well as investments in disruptive innovation. tage. It has also aimed to identify common patterns and
As per technology industry standards, such a challenges among the five national perspectives with
mix could include at least 20% investments in potentially greater relevance and applicability. Finally,
disruptive innovation, keeping in mind time- it has sought to contribute to ongoing efforts to shape
lines for technology maturation. Flexibility in defence-innovation systems by providing an empirically
implementation is important, though imple- driven analysis of innovation efforts in China, France,
mentation should ensure progressive oversight Germany, the UK and the US that highlights not just the
over innovation projects linked to technologi- more common quantitative metrics of defence innova-
cal maturity levels and rising costs, and avoid tion but brings out the criticality of qualitative metrics in
over-regulation in procurement. the assessment of innovation initiatives.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 29
Notes
1 Williamson Murray, ‘Clausewitz out, Computers in: Military 11 NATO Parliamentary Assembly, Defence Innovation Resolution

Culture and Technological Hubris’, The National Interest, 1 465, 23 November 2020, https://www.nato-pa.int/download-

June 1997. file?filename=/sites/default/files/2020-12/2020%20-%20

2 Jeremy Black, War and Technology (Bloomington, IN: Indiana NATO%20PA%20RESOLUTION%20465.pdf.

University Press, 2013), p. 27. 12 See, for example, Black, War and Technology. On the adaptation

3 McGregor Knox and Williamson Murray, The Dynamics of of civilian technologies in the development of lethal

Military Revolution, 1300–2050 (Cambridge: Cambridge autonomous weapon systems, see, for example, M.

University Press, 2001). Verbruggen, ‘The Role of Civilian Innovation in the

4 This is especially so if older technologies meet operational Development of Lethal Autonomous Weapon Systems’, Global

requirements, are robust and reliable, work better and can be Policy, vol. 10, no. 3, pp. 338–342.

readily standardised and replicated. The overall military, 13 Ben Barry, ‘Adapting in War’, Survival 54, no. 6, December

political and cultural context for the adoption of new 2012–January 2013, pp. 171–182.

technologies in defence is important, as new weapons systems 14 See Jon Schmid, ‘The Determinants of Military Technological

must be integrated in combined, joined arms and now multi- Innovation and Diffusion’, PhD dissertation, Georgia Institute

domain operations where they need to work cooperatively and of Technology, May 2018.

interoperably with legacy systems. Countries who rely mainly 15 An excellent analysis of the roots and nature of the revolution

on analogue capabilities will find it challenging to integrate AI, in military affairs is in Knox and Murray, The Dynamics of

cloud and edge computing in their military capabilities. Military Revolution, 1300–2050.

Equally, deploying these new technologies will be delayed if 16 Kikiras, ‘Disruptive Defence Innovations Ahead!’, p. 8.

their adoption entails larger scale changes in the operation of 17 Peter Dombrowski and Eugene Gholz, ‘Identifying Disruptive

military units, if it renders existing legacy platforms obsolete Innovation: Innovation Theory and the Defense Industry’,
faster or if the new technologies are not considered and proven Innovations: Technology, Governance, Globalization, vol. 4, no. 2,
reliable and robust enough for operational use. Spring 2009, pp. 101–118.

5 T. M. Cheung, T. G. Mahnken, and A.L. Ross, ‘Frameworks for 18 Anthony C. Pfaff, ‘The ethics of acquiring disruptive military

Analyzing Chinese Defense and Military Innovation’ in T. M. technologies’, Texas National Security Review, vol. 3, no. 1,

(Ed), Forging China’s Military Might: A New Framework for Winter 2019/2020, p. 41.

Assessing Innovation (Baltimore, MD: Johns Hopkins 19 Robert O. Work, ‘The United States − keeping the defence

University Press, 2014), pp. 15-46. innovation edge?’, IISS Defence Innovation Talks, 16 June 2021,

6 T. M. Cheung, ‘A conceptual framework of defence https://www.iiss.org/events/2021/06/us-defence-innovation-edge.

innovation’, Journal of Strategic Studies, 22 June 2021, pp. 1–27. 20 Emanuel Chiva, ‘A French perspective on defence innovation’,

7 D. P. Kikiras, ‘Disruptive Defence Innovations Ahead!’, European IISS Defence Innovation Talks, 29 April 2021, https://www.iiss.

Defence Matters, vol. 14, 2017, pp. 8–11, https://eda.europa.eu/ org/events/2021/04/defence-innovation-french-perspective.

webzine/issue14/cover-story/disruptive-defence-innovations-ahead. 21 Williamson Murray, Military Adaptation and War: With Fear of

8 M Wilkinson and S. Jewell, ‘UK Defence Innovation: design and Change (Cambridge: Cambridge University Press, 2012).

implementation of a system to realise value through exploitation 22 Jeremy Black, Military Strategy: A World History (New Haven,

of novelty’, Niteworks, 2017, p. 5, https://assets.publishing.service. CT: Yale University Press, 2020).

gov.uk/government/uploads/system/uploads/attachment_data/ 23 Eugene Gholz and Harvey M. Sapolsky,  ‘The defense

file/692445/NWP_-_UK_Defence_Innovation.pdf. innovation machine: Why the U.S. will remain on the cutting

9 Rob Murray, ‘Building a resilient innovation pipeline for the edge’,  Journal of Strategic Studies,  24 June 2021,

Alliance’, NATO Review, 1 September 2020, https://www.nato. DOI: 10.1080/01402390.2021.1917392.

int/docu/review/articles/2020/09/01/building-a-resilient- 24 Stephen Rodriquez, ‘How do you bridge the ‘valley of death’?

innovation-pipeline-for-the-alliance/index.html. Bring a ladder’ DefenseNews, 8 October 2020, https://www.

10 William B. Bonvillian and Charles Weiss, Technological Innovation d e f e n s e n e ws . co m/ o p i n i o n / co mme n t a ry / 2 0 20/ 10/ 08/

in Legacy Sectors, (Oxford: Oxford University Press, 2020). how-do-you-bridge-the-valley-of-death-bring-a-ladder/.

30 The International Institute for Strategic Studies

25 US Congress, ‘The National Security Commission on Artificial Gholz, Buying Military Transformation: Technological Innovation

Intelligence Final Report,’ April 2021, https://www.nscai.gov/ and the Defense Industry (New York, NY: Columbia University

wp-content/uploads/2021/03/Full-Report-Digital-1.pdf. Press, 2006).

26 Simona R. Soare, ‘Digital divide? Transatlantic defence 32 Capability gaps are recurrently identified as a driver of

cooperation on artificial intelligence,’ EUISS Policy Brief, no. 3, 5 defence innovation in a European/EU context. See Douglas

March 2020, https://www.iss.europa.eu/content/digital-divide- Barrie et al., ‘Defending Europe: scenario-based capability

transatlantic-defence-cooperation-ai; Graham Allison and Eric requirements for NATO’s European members’, IISS Research

Schmidt, ‘Is China Beating the U.S. to AI Supremacy?’, Harvard Paper, 10 May 2019, https://www.iiss.org/blogs/research-

University Belfer Center for Science and International Affairs Paper, paper/2019/05/defending-europe, and Simona R. Soare,

August 2020, https://www.belfercenter.org/sites/default/ ‘European Military AI: Why Regional Approaches are Lagging

files/2020-08/AISupremacy.pdf; Daniel Castro, Michael Behind’, in M. Raska and Z. Stanley-Lockman (eds.), Global

McLaughlin and Eline Chivot, ‘Who Is Winning the AI Race: Strategic Perspectives on Military AI (Singapore: Routledge,

China, the EU or the United States?’, Center for Data Innovation, forthcoming 2022).

19 August 2019, https://datainnovation.org/2019/08/ 33 For a comprehensive analysis of different models of

who-is-winning-the-ai-race-china-the-eu-or-the-united-states/. innovation see Bonvillian and Weiss, Technological Innovation

27 Exceptions are in William B. Bonvillian, ‘DARPA and its in Legacy Sectors. Examples of innovation models include open

ARPA-E and IARPA clones: a unique innovation organization and closed innovation; innovation pipelines; extended

model’, Industrial and corporate change, vol. 27, no. 5, October innovation pipelines; induced innovation; manufacturing-led

2018, pp. 897–914; Schmid, ‘The Determinants of Military innovation; innovation organisation; and innovation orchards.

Technological Innovation and Diffusion’; Black, War and For analyses on open and closed innovation systems see, for

Technology; and Knox and Murray, The Dynamics of Military example, Zoe Stanley-Lockman, ‘From closed to open systems:

Revolution, 1300–2050. How the US military services pursue innovation’, Journal of

28 The literature on defence innovation often portrays it by Strategic Studies, vol. 44, no. 4, 13 May 2021, https://doi.

reference to the adoption of EDTs in defence or the change in org/10.1080/01402390.2021.1917393.

procurement practices that is rapid or ‘at the pace of relevance’ 34 Dima Adamsky, The Culture of Military Innovation: The impact of

(emphasis added). However, it is worth noting that this cultural factors on the revolution in military affairs in Russia, the US,

description is a misnomer since profound changes in defence and Israel (Stanford, CA: Stanford University Press, 2010), and

establishments historically are long-term endeavours regardless Theo Farrell, Sten Rynning and Terry Terriff, Transforming Military

of the complexity of the technologies adopted. While it may be Power since the Cold War: Britain, France, and the United States, 1991-

conceivable that the current iteration of defence innovation 2012 (Cambridge, UK: Cambridge University Press, 2013).

may be faster than previous ones, it is nevertheless important to 35 Tim Ripley, ‘UK to deploy more intelligence personnel to

manage expectations about how fast changes will materialize. monitor Russia and China’, Janes, 17 September 2021, https://

For an excellent analysis of historical trends in military customer.janes.com/Janes/Display/BSP_5712-JDW.

revolution, see Knox and Murray, The Dynamics of Military 36 Jeremy Black, Geopolitics and the Quest for Dominance

Revolution, 1300-2050, and of the role of cumulative innovation (Bloomington, IN: Indiana University Press, 2016), and

in defence, see Black, War and Technology. Williamson Murray, War, Strategy, and Military Effectiveness

29 Michael C. Horowitz and Shira Pindyck, ‘What Is a Military (Cambridge, UK: Cambridge University Press, 2011).

Innovation and Why It Matters’, SSNR, unpublished 37 Patrick M. Morgan and Erik Gartzke, Cross-Domain Deterrence:

manuscript, 23 December 2019, p. 3, https://papers.ssrn.com/ Strategy in an Era of Complexity (New York, NY: Oxford

sol3/papers.cfm?abstract_id=3504246. University Press, 2019).

30 Stephen Peter Rosen, Winning the Next War: Innovation and the 38 Schmid, ‘The Determinants of Military Technological

Modern Military (Ithaca, NY: Cornell University Press, 1991). Innovation and Diffusion’.

31 Barry R. Posen, The Sources of Military Doctrine: France, Britain, 39 Pfaff, ‘The ethics of acquiring disruptive military technologies’.

and Germany between the World Wars (Ithaca, NY: Cornell 40 American expert, GMF Transatlantic Task Force on

University Press, 1984), and Peter Dombrowski and Eugene Technology and Security, 12 October 2021.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 31
41 Work, ‘The United States − keeping the defence innovation edge?’. envoy’, Reuters, 4 February 2021, https://www.reuters.com/

42 Dstl, ‘Game on: Inside the Dstl’s Defence Wargaming Centre’, article/us-usa-russia-newstart-idUSKBN2A417F.

retrieved 29 September 2021, https://defence.nridigital.com/ 53 Michael R. Pompeo, ‘Speech by Secretary of State Michael R.

global_defence_technology_mar20/game_on_inside_dstls_ Pompeo at the Arctic Council Ministerial Meeting’, US

defence_wargaming_centre. Department of State, 7 May 2019, https://oaarchive.arctic-council.

43 France, Ministère des Armées, ‘Découvrir la Red Team’, org/bitstream/handle/11374/2409/2019_Rovaniemi_Ministerial_

retrieved 29 September 2021, https://redteamdefense.org/ Statement_by_the_USA.pdf?sequence=1&isAllowed=y.

decouvrir-la-red-team. 54 Work, ‘The United States − keeping the defence innovation

44 Bundeswehr, ‘Resilient Response 2020 Successfully edge?’. Also see John R. Allen, Ben Hodges and Julian

Completed’, 7 December 2020, https://www.bundeswehr.de/ Lindsey-French, Future War and the Defence of Europe (Oxford:

en/organization/bundeswehr-medical-service-/resilient-response- Oxford University Press, 2021).

2020-successfully-completed--4882638. 55 The State Council Information Office of the People’s Republic

45 Riccardo Cociani,  ‘China’s military-civil fusion aids the of China, ‘White Paper on China’s National Defence in the

growth of wargames’, Janes, 27 July 2021. New Era’, Xinhua, 24 July 2019, http://www.xinhuanet.com/

46 For an in-depth analysis of defence innovation and European english/2019-07/24/c_138253389.htm.

strategic autonomy, see Simona R. Soare, ‘Algorithmic power? 56 Elsa B. Kania, ‘Is China lagging behind the military-

Artificial intelligence and European strategic autonomy’, in D. technological race?’ IISS Defence Innovation Talks, 7 July

Broeders et al., Artificial Intelligence and International Conflict in 2021, https://www.iiss.org/events/2021/07/china-defence-innovation.

Cyberspace (London: Routledge, forthcoming 2021). For a 57 See, for example, Elsa Kania and Mauro Gilli, ‘Is China

comparative analysis of multilateral approaches to defence lagging behind the military-technological race?’.
innovation in the EU and NATO, see Soare, ‘European 58 Meia Nouwens, ‘Is China lagging behind the military-

Military AI: Why Regional Approaches are Lagging Behind’. technological race?’.

47 Schmid, ‘The Determinants of Military Technological 59 For a historical overview of this phenomenon in other instances of

Innovation and Diffusion’. revolutions in military affairs see Black, War and Technology, p. 5.

48 Interview with German MoD official, September 2021. 60 US Department of Defense, ‘Military and Security

49 There is significant historical continuity in both American and Developments Involving the People’s Republic of China 2020:

Chinese approaches towards defence innovation, notably the Annual Report to Congress’, 1 September 2020, https://media.

prominence of American techno-centric and technophile defense.gov/2020/Sep/01/2002488689/-1/-1/1/2020-DOD-

strategic culture, the Chinese complex about overcoming its CHINA-MILITARY-POWER-REPORT-FINAL.PDF.

generational technological gaps and the mutual fear of 61 Elsa B. Kania, ‘“AI weapons” in China’s military innovation’,

technological surprise as a cause of military defeat as well as Brookings Institution, April 2020, https://www.brookings.edu/

the loss of strategic competitiveness. research/ai-weapons-in-chinas-military-innovation/, and Elsa

50 Steven Erlanger and Michael D. Shear, ‘Shifting Focus, NATO B. Kania, ‘Artificial intelligence in China’s revolution in

Views China as a Global Security Challenge’, New York Times, military affairs’, Journal of Strategic Studies, vol. 44, no. 4, 25

14 June 2021, https://www.nytimes.com/2021/06/14/world/ May 2021, https://doi.org/10.1080/01402390.2021.1894136.

europe/biden-nato-china-russia.html. 62 Ryan Fedasiuk, ‘Chinese Perspectives on AI and Future

51 Julian E. Barnes and Adam Satariano, ‘U.S. Campaign to Ban Military Capabilities’, CSET, August 2020, https://cset.

Huawei Overseas Stumbles as Allies Resist’, The New York georgetown.edu/publication/chinese-perspectives-on-ai-and-

Times, 17 March 2019, https://www.nytimes.com/2019/03/17/ future-military-capabilities/, and Kevin Pollpeter and

us/politics/huawei-ban.html. Amanda Kerrigan, ‘The PLA and Intelligent Warfare: A

52 Jack Detsch, ‘Trump Wants China on Board With New Arms Preliminary Analysis’, Center for Naval Analyses (CNA), 4

Control Pact’, Foreign Policy, 23 July 2020, https:// October 2021, https://www.cna.org/CNA_files/PDF/The-PLA-

foreignpolicy.com/2020/07/23/trump-china-russia-new- and-Intelligent-Warfare-A-Preliminary-Analysis.pdf.

arms-control-agreement-start/, and Stephanie Nebehay, 63 The concept of military dominance remains present in

‘Biden to pursue arms control, seeks to engage China: U.S. American military documents and narratives, even as a part of

32 The International Institute for Strategic Studies

 its defence establishment is increasingly moving towards an 72 Ibid.

understanding that overarching military dominance over other 73 British MoD official, DSEI: Powering Progress, Defining your

great powers may no longer be possible and that the US will Future, 12–15 September 2021, https://www.dsei.co.uk.

increasingly operate in an environment of military technological 74 ‘Attritable’ capabilities signal a new focus on combat mass and

near parity or sectoral parity. For more on this latter approach, attrition tactics. In military jargon, the concept refers to the use

see US Department of Defense, the Joint Staff, ‘Description of of an increased number of sophisticated yet cheaper

the National Military Strategy 2018’, https://www.jcs.mil/ unmanned military platforms which can be deployed in

Portals/36/Documents/Publications/UNCLASS_2018_ multiple operations, even at the cost of losing them if the

National_Military_Strategy_Description.pdf. tactical or operational gain is worth it. For an account of

64 Robert O. Work, remarks at the NCASI Global Emerging attritable aircraft, see Sidney Dean, ‘Attritable Aircraft’,

Technologies Summit, 13 July 2021, https://www.nscai.gov/ European Security and Defence, 31 August 2021, https://euro-sd.

all-events/summit/. com/2021/08/articles/exclusive/23250/attritable-aircraft/..

65 US Department of Commerce, Gina Raimondo, remarks at the 75 General Carter, ‘Defence Innovation Talks: a conversation

NSCAI Global Emerging Technologies Summit, 13 July 2021, with General Sir Nick Carter’.

https://www.nscai.gov/all-events/summit/. 76 British MoD official, DSEI: Powering Progress, Defining your Future.

66 Ministère des Armées, Strategic Update 2021, February 2021, 77 Atos, ‘The French Defense Procurement Agency selects the

https://sites.breakingmedia.com/uploads/sites/3/2021/04/ consortium led by Atos for Project Artemis, phase II’, 23

strategic-update-2021.pdf. March 2019, https://atos.net/en/2019/press-release_2019_05_23/

67 Official with Ministère des Affaires étrangères, GMF the-french-defense-procurement-agency-selects-the-

Transatlantic Task Force on Technology and Security, 12 consortium-led-by-atos-for-project-artemis-phase-ii#_ftn1.

October 2021. 78 Ministère des armées, ‘Vulcain : à la découverte de la future

68 For British and French military commanders, recent strategic section robotique’, 29 June 2021, https://www.defense.gouv.fr/terre/

developments present important lessons for the future of actu-terre/vulcain-a-la-decouverte-de-la-future-section-robotique.

armed conflict – particularly regarding assumptions of mass, 79 UK Army, ‘Artificial Intelligence used on British Army

concentration of power and detection. The rapid Russian force operation for the first time’, 5 July 2021, https://www.army.

concentration and dispersion on the Ukrainian border in 2020, mod.uk/news-and-events/news/2021/07/artificial-intelligence-

the Chinese PLA’s air and naval manoeuvres in the Taiwan used-on-british-army-operation-for-the-first-time/.

Strait, the use of uninhabited systems in the 2021 Nagorno- 80 Harry Lee, ‘Machine learning for the future fleet: Dstl’s

Karabakh War and the Russian uses of AI-enabled tactical Intelligent Ship’, Global Defence Technology, February 2021,

command and control of inhabited and uninhabited systems https://defence.nridigital.com/global_defence_technology_

in Syria and Libya are seen to hold important lessons. feb21/ai_dstl_intelligent_ship.

69 Barrie et al., ‘European defence policy in an era of renewed 81 Chiva, ‘A French perspective on defence innovation’.

great-power competition’, IISS Research Paper, 17 February 82 Ministère des Affaires étrangères official, GMF Transatlantic

2020, p. 2, https://www.iiss.org/blogs/research-paper/2020/02/ Task Force on Technology and Security, 12 October 2021.

the-future-of-european-defence. 83 Interview with German MoD official, September 2021.

70 Dominic Nicholls, ‘British Army numbers cut to smallest in 84 Interview with German MoD official, October 2021.

history as revamp revealed in Defence Review’, The Telegraph, 85 Ibid.

23 March 2021, https://www.telegraph.co.uk/news/2021/03/22/ 86 Ibid.

watch-livearmy-numbers-cut-smallest-history-overhaul- 87 Ibid.

armed-forces/. 88 German Federal Ministry of Defence, ‘Key Elements of the

71 General Sir Nick Carter, ‘Defence Innovation Talks: a Bundeswehr of the Future’, 18 May 2021, https://www.bmvg.

conversation with General Sir Nick Carter’, and Interview de/resource/blob/5092636/f30cf5dd345488be5c5a090c58367fae/

with German MoD official, IISS Defence Innovation Talks, 30 eckpunkte-final-en-data.pdf.

March 2021, https://www.iiss.org/events/2021/03/ 89 Interview with German MoD official, September 2021.

defence-innovation-talks-general-sir-nick-carter. 90 Interview with German MoD official, October 2021.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 33
91 S. Sprenger, ‘German defense minister vows to keep fighting 102 Xi Jinping apud Kania, ‘“AI weapons” in China’s military

for armed drones’, DefenseNews, 16 April 2021, https://www. innovation’.

defensenews.com/global/europe/2021/04/16/german-defense- 103 Nouwens, ‘Is China lagging behind the military-

secretary-vows-to-keep-fighting-for-armed-drones/. technological race?’.

92 Interviews with German MoD officials, September and 104 Work, ‘The United States − keeping the defence innovation

October 2021. edge?’.

93 Among other states, France has raised concerns that when it 105 Katrina Manson, ‘US has already lost AI fight to China, says

comes to future autonomous and AI-driven weapon systems, ex-Pentagon software chief’, Financial Times, 10 October 2021,

Europeans are not well placed to impose/enforce international https://www.ft.com/content/f939db9a-40af-4bd1-b67d-10492

ethical norms of use and stringent regulation may only serve 535f8e0.

to give away European advantages to adversaries in the global 106 Work, ‘The United States − keeping the defence innovation

competition for power. French official with the Ministère des edge?’.

Affaires étrangères, GMF Transatlantic Task Force on 107 Empirical evidence suggests a mix of top-down and bottom-up

Technology and Security, 12 October 2021. innovation is needed and mature defence innovation

94 Interviews with German MoD officials, September and ecosystems like the US are built on this principle of

October 2021. complementarity.

95 Sharon Weinberger, The Imagineers of War: The Untold Story of 108 JAIC seeks to streamline EDTs into innovation efforts across

DARPA, the Pentagon Agency That Changed the World (New York, the defence enterprise and coordinate innovation efforts. DIU

NY: Alfred A. Knopf, 2017), and W.B. Bonvillian, ‘DARPA and is responsible for outreach and broader engagement with

its ARPA-E and IARPA clones’, Industrial and Corporate Change, private industry and start-ups. SCO is tasked with delivering

vol. 27, no. 5, 2018, pp. 897–914, doi: 10.1093/icc/dty026. innovative solutions for short-term operational needs. And

96 Simona R. Soare, ‘Politics in the machine: The political context DARPA explores more foundational innovation challenges,

of emerging technologies, national security, and great power such as third-wave AI technologies.

competition’, in Reuben Steff, Joe Burton and Simona R. Soare 109 Bonvillian, ‘DARPA and its ARPA-E and IARPA clones’.

(eds.), Emerging Technologies and International Security: Machines, 110 President Emmanuel Macron, ‘Initiative pour l’Europe -

the State, and War (London: Routledge, 2020), pp. 103–122. Discours d’Emmanuel Macron pour une Europe souveraine,

97 Laura Silver, Kat Devlin and Christine Huang, ‘Most Americans unie, démocratique’, Palais Elysée, 26 September 2017, https://

Support Tough Stance Toward China on Human Rights, www.elysee.fr/emmanuel-macron/2017/09/26/initiative-pour-

Economic Issues’, PEW Research Center, 4 March 2021, https:// l-europe-discours-d-emmanuel-macron-pour-une-europe-

www.pewresearch.org/global/2021/03/04/most-americans-support- souveraine-unie-democratique. President Macron’s proposal

tough-stance-toward-china-on-human-rights-economic-issues/. for the establishment of an EU Innovation Agency was

98 Soare, ‘European Military AI: Why Regional Approaches are rejected by other member states and the commission. Instead,

Lagging Behind’. political consensus led to the establishment of the economy-

99 Interview with German MoD official, October 2021. focused European Innovation Council, under the European

100 Examples include the US Department of Defence, Summary of Commission’s remit.

the 2018 Department of Defence Artificial Intelligence Strategy: 111 General Secretariat of the Council, ‘Special meeting of the

Harnessing AI to Advance Our Security and Prosperity, 12 European Council (24 and 25 May 2021) – Conclusions’, 24–25

February 2019, https://media.defense.gov/2019/Feb/12/200208 May 2021, https://www.consilium.europa.eu/media/49791/2425-

8963/-1/-1/1/SUMMARY-OF-DOD-AI-STRATEGY.PDF. 05-21-euco-conclusions-en.pdf.

101 Notable examples are the United States’ Joint Artificial 112 German Federal Ministry of Defence, ‘Gaining technology

Intelligence Centre (JAIC) and the Defence Innovation Unit sovereignty - the new cyber agency’, 24 September 2021,

(DIU), the UK’s Defence Science and Technology Laboratory https://www.bmvg.de/de/aktuelles/technologiesouveraenitaet-

(Dstl), Defence AI Centre and new Technology Accelerator, erlangen-die-neue-cyberagentur-27996.

France’s Agence de l’innovation de défense and Germany’s 113 Interview with German MoD official, October 2021.

Cyber Innovation Hub. 114 Ibid.

34 The International Institute for Strategic Studies

115 German Army Concepts and Capabilities Development 131 Aaron Mehta, ‘UK Signs Up For Anduril AI Base Defense

Centre, ‘Artificial Intelligence in Land Forces Position’, 2019, Technology’, DefenseNews, 14 September 2021, https://

https://www.bundeswehr.de/resource/blob/156026/79046a243 breakingdefense.com/2021/09/uk-signs-up-for-anduril-ai-base-

22feb96b2d8cce168315249/download-positionspapier- defense-technology/.

englische-version-data.pdf. 132 See, for example, Behörden Spiegel, ‘Defence Innovation Talk

116 Chiva, ‘A French perspective on defence innovation’. 5 – Future Technologies for the Bundeswehr’, 8 October 2020,

117 Ibid. https://www.youtube.com/watch?v=TxsJxeXb_B0.

118 Agence de l’innovation de défense, ‘Document de Référence 133 Elaine L. Halchin, ‘Other Transaction (OT) Authority’,

de l’Orientation de l’Innovation de Défense DROID 2021’, p. Congressional Research Service Report RL34760, 15 July 2011,

24, https://www.defense.gouv.fr/aid. https://sgp.fas.org/crs/misc/RL34760.pdf.

119 Airbus and Fraunhofer Institutes, ‘The responsible use of new 134 Defense Innovation Board, ‘Technology and Capabilities -

technologies in a Future Combat Air System (FCAS)’, retrieved Recommendation 6: Expand Use of Available Acquisition

29 September 2021, https://www.fcas-forum.eu/en. Waivers and Exemptions’, January 2017, https://innovation.

120 Weinberger, The Imagineers of War, and Bonvillian, ‘DARPA defense.gov/Recommendations/.

and its ARPA-E and IARPA clones’. 135 Chiva, ‘A French perspective on defence innovation’. Also see

121 Defense Innovation Board, ‘Practices and Operations - Soare, ‘European Military AI: Why Regional Approaches are

Recommendation 13: Design a DoD Fast-Track for Major Lagging Behind’.

Technology Initiatives’, January 2017, https://innovation. 136 General Carter, ‘Defence Innovation Talks: a conversation

defense.gov/Recommendations/. with General Sir Nick Carter’.

122 Agence de l’innovation de défense, ‘Innovation Défense Lab: 137 Chiva, ‘A French perspective on defence innovation’.

Un accélérateur d’Innovation pour l’Ensemble de la Défense’, 138 Airbus, ‘Digital Design, Manufacturing & Services: Transforming

23 September 2021, https://www.defense.gouv.fr/aid/l-agence/ Airbus through digital continuity’, 4 January 2021, https://www.

innovation-defense-lab. airbus.com/innovation/industry-4-0/digital-design-and-

123 Chiva, ‘A French perspective on defence innovation’. manufacturing-ddms.html.

124 For insights into some of the known processes, see Elsa B. Kania 139 Interview with German MoD official, September 2021.

and Lorand Laskai, ‘Myths and Realities of China’s Military- 140 S. Rodriguez, ‘Top 10 Failed Defence Programs of the RMA

Civil Fusion Strategy’, Center for New American Security Era’, War on the Rocks, 2 December 2014, https://warontherocks.

(CNAS), 28 January 2021, https://s3.us-east-1.amazonaws.com/ com/2014/12/top-10-failed-defense-programs-of-the-rma-era/.

files.cnas.org/documents/Myths-and-Realities-of- 141 Marcello E. Gallo, ‘Defense Advanced Research Projects

China%E2%80%99s-Military-Civil-Fusion-Strategy_FINAL- Agency: Overview and Issues for Congress’, Congressional

min.pdf?mtime=20210127133521&focal=none. Research Service Report R45088, 19 August 2021, https://sgp.

125 Nouwens, ‘Is China lagging behind the military-technological fas.org/crs/natsec/R45088.pdf.

race?’. 142 European Defence Agency, ‘Defence Data 2018-2019: Key

126 Kania, ‘Is China lagging behind the military-technological race?’. Findings and Analysis’, 28 January 2021, https://eda.europa.

127 Nouwens, ‘Is China lagging behind the military-technological eu/docs/default-source/brochures/2019-eda-defence-data-

race?’. report.pdf, and NATO, ‘Defence Expenditure of NATO

128 Ibid. Countries (2014-2021)’, Press Communique PR/CP(2021)094,

129 According to Nouwens, such practices include investment in 11 June 2021, https://www.nato.int/nato_static_fl2014/assets/

early-stage foreign companies and start-ups that are small pdf/2021/6/pdf/210611-pr-2021-094-en.pdf.

enough in value to fly under the radar of FDI screening 143 Bonvillian, ‘DARPA and its ARPA-E and IARPA clones’.

mechanisms and be publicly scrutinised. Ibid. 144 Weinberger, The Imagineers of War.

130 For an analysis challenging the mainstream argument that 145 Defense Advanced Research Projects Agency, ‘Assault Breaker’,

defence establishments lack sufficient access to private industry retrieved 29 September 2021, https://www.darpa.mil/about-us/

technological development, see Gholz and Sapolsky, ‘The timeline/assault-breaker#:~:text=In%201978%2C%20DARPA%20

defense innovation machine’. integrated%20a,under%20its%20Assault%20Breaker%20program.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 35
146 United States Government Accountability Office, ‘Report to de l’Orientation de l’Innovation de Défense DROID 2021’, p. 9,

Congressional Committees on Defence Science and Technology: https://www.defense.gouv.fr/aid.

Adopting Best Practices Can Improve Innovation Investments 159 Cour de comptes, ‘Mission Défense: Note d’analyse de

and Management’, June 2017, https://www.gao.gov/assets/gao- l’exécution budgétaire 2020’, April 2020, p. 56, https://www.

17-499.pdf. ccomptes.fr/sites/default/files/2021-04/NEB-2020-Defense.pdf.

147 General Carter, ‘Defence Innovation Talks: a conversation 160 Laurent Saint Martin et François Cornu-Gentille, ‘Rapport fait

with General Sir Nick Carter’, and Work, ‘The United States − au nom de la Commission des Finances, de l’Économie

keeping the defence innovation edge?’. Générale et du Contrôle Budgétaire sur le Project de Loi, après

148 European Commission official, GMF Transatlantic Task Force engagement de la procédure accélérée de règlement du

on Technology and Security, 12 October 2021. budget et d’approbation des comptes de l’année 2020 (n°

149 Work, ‘The United States − keeping the defence innovation 4090). Annexe n° 13, Défense : Préparer pour l’Avenir’,

edge?’. Assemblée Nationale, 26 May 2021, pp. 28–32, https://www.

150 Andrea Asoni, Andrea Gilli, Mauro Gilli and Tino Sanandaji, assemblee-nationale.fr/dyn/15/rapports/cion_fin/

‘Skill-biased technological change and the demographic l15b4195-a13_rapport-fond.pdf.

composition of the US military, 1979–2008’, VoxEU, 19 161 Interview with German MoD official, October 2021.

September2021,https://voxeu.org/article/skill-biased-technological- 162 N. Lomas, ‘UK to invest in AI and cyber as part of major

change-and-demographic-composition-us-military. defense spending hike’, TechCrunch, 19 November 2020.

151 Nouwens, ‘Is China lagging behind the military-technological 163 Ministry of Defence of the United Kingdom Press Release, ‘Army

race?’. robotics receive £66-million boost’, 5 March 2019, https://www.

152 An excellent analysis of China’s multiple talent acquisition gov.uk/government/news/army-robotics-receive-66m-boost.

programmes and initiatives is in Wm. C. Hannas and Huey- 164 Aerospace and Defence Industries Association of Europe,

meei Chang, ‘China’s Access to Foreign AI Technology: An ‘The Aerospace and Defence Industries Association of Europe:

Assessment’, CSET, September 2019, https://cset.georgetown. 2020 Facts & Figures’, 2020, https://www.asd-europe.org/

edu/wp-content/uploads/CSET_China_Access_To_Foreign_ sites/default/files/atoms/files/ASD_FactsFigures_2020.pdf.

AI_Technology-1.pdf. 165 R. Berger, ‘The road to AI – Investment dynamics in the

153 Gallo, ‘Defense Advanced Research Projects Agency’, and European ecosystem. AI Global Index 2019’, Roland Berger

Gholz and Sapolsky, ‘The defense innovation machine’. website, 7 January 2020, https://www.rolandberger.com/en/

154 US Congress, National Defence Authorisation Acts, Insights/Publications/The-road-to-AI.html.

2018–2022. 166 Work, ‘The United States − keeping the defence innovation edge?’.

155 IISS, Military Balance + Database, retrieved 4 October 2021, 167 Ibid.

https://www.iiss.org/publications/the-military-balance-plus. 168 Gallo, ‘Defense Advanced Research Projects Agency’.

For an estimate of Chinese R&D spending, see Ashwin 169 Defence Science and Technology Laboratory (Dstl), ‘Annual

Acharya and Zachary Arnold, ‘Chinese Public AI R&D Report and Accounts 2019/2020’, 2021, https://assets.publishing.

Spending: Provisional Findings’, CSET Analysis, December service.gov.uk/government/uploads/system/uploads/

2019, https://cset.georgetown.edu/publication/chinese-public- attachment_data/file/923884/20201001-Dstl_ARAC_2019-20_

ai-rd-spending-provisional-findings/. FINAL_v1_0-O_WEB-OPTIMISED.pdf
156 European Defence Agency, ’Defence data 2018-2019: Key findings 170 Agence de l’innovation de défense, ‘Document de Référence

and analysis’, 28 January 2021, p. 10, https://eda.europa.eu/docs/ de l’Orientation de l’Innovation de Défense DROID 2021’.

default-source/brochures/2019-eda-defence-data-report.pdf. 171 United States Government Accountability Office, ‘Report to

157 Ministère des Armées, ‘Discours de Florence Parly, ministre Congressional Committees on Defence Science and

des Armées : Intelligence artificielle et défense’, 5 April 2019, Technology’.

https://www.defense.gouv.fr/salle-de-presse/discours/ 172 Soare, ‘European Military AI: Why Regional Approaches are

discours-de-florence-parly/discours-de-florence-parly-ministre- Lagging Behind’.

des-armees_intelligence-artificielle-et-defense. 173 Agence de l’innovation de défense, ‘Document de Référence

158 Agence de l’innovation de défense, ‘Document de Référence de l’Orientation de l’Innovation de Défense DROID 2021’.

36 The International Institute for Strategic Studies

174 Halo effects refer to the influence that government-driven investors interested in supporting start-ups that commercialise

areas of research exert on venture capital markets and private new technologies.

Acknowledgements

The authors would like to acknowledge a Microsoft grant that made possible the 2021 IISS Defence Innovation Talks series on
which this analysis builds. They are grateful to all the governmental and academic experts who generously donated their time and
insights to support this analysis. They wish to thank Dr Bastian Giegerich for his review of and comments on earlier drafts of this
paper, and are also grateful for the research assistance provided by Fabian Hoffmann, Josephine Niehoff and Julian Wachtel. The
views expressed in this paper prepared by members of the IISS research staff do not necessarily reflect those of the participants
in the 2021 IISS Defence Innovation Talks, or the experts interviewed in the preparation of this publication.

Leading Edge: Key Drivers of Defence Innovation and the Future of Operational Advantage 37
© The International Institute for Strategic Studies, 2021

The International Institute for Strategic Studies – UK

Arundel House | 6 Temple Place | London | wc2r 2pg | UK
t. +44 (0) 20 7379 7676 f. +44 (0) 20 7836 3108 e. [email protected] www.iiss.org

The International Institute for Strategic Studies – Americas

2121 K Street, NW | Suite 600 | Washington, DC 20037 | USA
t. +1 202 659 1490 f. +1 202 659 1499 e. [email protected]

The International Institute for Strategic Studies – Asia

9 Raffles Place | #49-01 Republic Plaza | Singapore 048619
t. +65 6499 0055 f. +65 6499 0059 e. [email protected]

The International Institute for Strategic Studies – Europe

Pariser Platz 6A | 10117 Berlin | Germany
t. +49 30 311 99 300 e. [email protected]

The International Institute for Strategic Studies – Middle East

14th floor, GBCORP Tower | Bahrain Financial Harbour | Manama | Kingdom of Bahrain
t. +973 1718 1155 f. +973 1710 0155 e. [email protected]