FST JOURNAL

The author writes in a personal capacity. The views expressed are the author’s own and do not represent the institutional position of University College London.

UK research-intensive universities are well positioned to make substantial contributions to national security, but doing so effectively requires both the government and universities to navigate six structural tensions where legitimate institutional values conflict. Understanding these tensions enables both universities and government to develop frameworks for sustained, productive engagement.

Recent debates about research partnerships, student and staff concerns about defence industry links, and questions about export controls reveal the complexity universities face. These are not obstacles to engagement but governance challenges requiring thoughtful policy support.

Here, I focus on the relatively small number of universities that conduct research, innovation and commercialisation at scale across advanced technology domains. Perhaps ten to twenty exist in the UK, and a similar number across Europe. They possess world-leading capabilities in areas like artificial intelligence, quantum technologies, and advanced materials. They have precisely the expertise that government seeks to mobilise for defence innovation and broader national security through initiatives spanning UK Defence Innovation (UKDI), the Strategic Defence Review, and programmes including AUKUS Pillar II.

Across Europe, this subset of universities confronts similar pressures as defence budgets grow, strategic autonomy frameworks recognise research as foundational to security, and geopolitical competition demands technological advantage. The UK’s post-Brexit position creates distinctive opportunities, as many UK institutions have strong defence research traditions. All have flexibility from operating outside EU frameworks while maintaining European scientific partnerships. And the Prime Minister’s emphasis on national security as encompassing economic resilience, technological sovereignty, and societal preparedness creates natural alignment with the broader research strengths of UK universities.

This paper identifies six structural tensions underpinned by two fundamental ambiguities that require updated policy frameworks. Rather than prescriptive solutions, I seek to clarify where competing institutional values require balancing and why sustained engagement depends on governance and policy frameworks that acknowledge these complexities.

Universities, as independent institutions, can make principled choices about which values to prioritise. But doing so effectively requires both institutional clarity and policy support. The goal is enabling productive long-term partnerships by making trade-offs explicit and creating frameworks that allow different institutions to contribute according to their particular missions and capabilities.

Two Fundamental Ambiguities

 Two underlying ambiguities shape all discussion of national security research and understanding them is essential for developing productive engagement frameworks.

The erosion of civil-military boundaries 

It has become impossible to categorise research definitively as either “civil” or “military” in scope. This reflects two developments. First, modern security threats increasingly involve areas like cyber operations, information warfare, supply chain resilience, and societal preparedness rather than solely conventional military capabilities. Research on topics as diverse as social media algorithms, semiconductor supply chains, or pandemic preparedness can all strengthen national security without traditional weapons applications. A computer scientist studying misinformation propagation contributes simultaneously to democratic resilience and information security. Second, defence establishments now recognise that strategic advantage emerges from an extraordinarily broad knowledge base and its applied uses encompassing disciplines such as behavioural psychology, climate science, urban planning, and biosecurity alongside traditional defence domains.

This breadth means the full strength of UK research-intensive universities becomes relevant to national security. The same algorithms serve autonomous vehicles and autonomous systems. The same quantum sensors enable medical imaging and advanced detection. The same materials science improves energy systems and protective technologies. The fundamental capabilities of research-intensive universities can directly support national security without requiring distinct “military” programmes. The Strategic Defence Review’s emphasis on emerging technologies, MoD partnerships in AI and quantum, and programmes like UKDI all recognise domains where research-intensive universities possess world-leading capabilities. But governance frameworks that depend on clear civil-military distinctions cannot accommodate this convergence. When quantum computing, AI, and advanced materials inherently serve both civilian and security ends, categorical prohibitions become incoherent.

 The fluidity of strategic relationships

Determining appropriate partners for research with national security relevance requires more nuanced judgment than Cold War friend-adversary distinctions allowed. China is now simultaneously a major research partner, economic competitor, strategic rival, and security concern. The 2025 Strategic Defence Review groups Russia, Iran and North Korea as threats requiring the UK to move to warfighting readiness alongside describing China as ‘a sophisticated and persistent challenge’, a significantly harder framing than the previous Integrated Review’s characterisation of China as a “systemic challenge”. Russia and Iran are subject to comprehensive sanctions making collaboration straightforwardly prohibited. Ambiguity also lies elsewhere. Middle Eastern states, Turkey, and numerous other countries resist simple categorisation as partners or adversaries. Some partnerships are therefore clearly prohibited, while others are clearly permitted with trusted allies. But a substantial middle ground requires sophisticated judgment about which collaborations serve UK interests versus creating genuine security risks.

 A university collaborating with institutions in a competitor state on quantum computing must balance maintaining scientific leadership against potential technology transfer risks. Export control regimes provide one framework for managing these complexities, though their application to emerging technologies requires ongoing refinement. The UK government’s Trusted Research agenda and Research Collaboration Advice Team (RCAT) represent progress toward supporting universities in navigating partnership risks. But universities need clearer policy signals about which partnerships government considers strategically valuable versus problematic, along with support for the due diligence infrastructure required to make sound judgments.

 These ambiguities compound each other

When research serves both civilian and security purposes, and when international partnerships resist categorical classification, traditional frameworks permitting “peaceful research with allied partners” while prohibiting “weapons development with rivals” become inoperable. Nearly all research partnerships with nearly all countries become contestable. Universities must develop case-by-case judgment about strategic implications, geopolitical risks, and dual-use potential for significant collaborations. This forces a level of security assessment that some institutions currently lack the expertise to conduct. The gap is not one of willingness but of capacity, requiring investment in specialised knowledge and infrastructure. Government can enable university contribution to national security by providing clear demand signals, supporting development of institutional capacity for strategic assessment, and creating frameworks that acknowledge inherent complexity rather than demanding impossible categorical clarity.

Six Structural Tensions

 Against this backdrop, research-intensive universities must navigate six structural tensions where legitimate institutional values conflict. Different institutions will balance these differently based on their traditions, national contexts, and missions. The value for both universities and government comes from understanding what is genuinely at stake.

Individual versus institutional academic freedom

The freedom of individual researchers to pursue knowledge can conflict with institutional autonomy to shape collective identity. When a researcher claims academic freedom to accept defence funding while the institution claims academic freedom to decline such positioning, both make legitimate appeals to the same principle but mean different things. In Germany, constitutional protections for Wissenschaftsfreiheit create high bars for institutional restrictions on individual research choices. In the UK, academic freedom is a professional norm, giving institutions more flexibility but less principled grounding for constraints. The erosion of civil-military boundaries amplifies this tension. When a computer scientist can reasonably argue that their machine learning research serves both civilian and security purposes, institutions must either make contentious judgments about ultimate use or apply restrictions so broadly that they constrain significant swathes of legitimate science. Frameworks that evaluate research based on substantive criteria such as openness, publication, and collaborative relationships are more defensible than categorical prohibitions based on funding source alone.

 Clean hands versus critical engagement

Universities adopt different postures toward security research based on their histories and national contexts. This diversity can be a strength. Some institutions emphasise maintaining analytical distance, arguing that intellectual independence enables more valuable long-term contributions to sound security policy. They study defence technologies, analyse strategic implications, and educate future security professionals without direct involvement in capability development. Others emphasise active engagement, arguing that universities have a responsibility to embed ethical frameworks into defence technology development. If autonomous systems and AI will be developed regardless, participation enables shaping how technologies evolve in ways serving democratic values and international humanitarian law. A third approach, emerging from Nordic “total defence” frameworks, positions universities as contributing to societal resilience alongside other sectors. This neither emphasises moral distance nor frames engagement primarily as weapons development, but rather as fulfilling societal responsibilities for comprehensive national security including cyber resilience, supply chain security, and technological sovereignty. This approach aligns naturally with the Strategic Defence Review’s emphasis on a ‘whole-of-society’ approach to national security. All three approaches can legitimately contribute to UK national security, and policy should enable this diversity.

National security versus international collaboration

Research-intensive universities derive their excellence from international collaboration and knowledge exchange. This itself strengthens UK research competitiveness and technological advantage. The question is not whether to maintain international partnerships but how to do so in ways that advance rather than compromise national security. Can universities maintain selective partnerships with institutions in strategic rival states while contributing to UK quantum capabilities? Can they balance security screening with attracting international talent that strengthens UK research capacity? These are navigable challenges requiring appropriate frameworks rather than binary categorisation. The complexity of contemporary strategic relationships also demands sophistication. For example, is collaboration with specific institutions on quantum research acceptable if framed as fundamental science with appropriate safeguards; what due diligence on partner institution links is sufficient; and which technical domains require enhanced scrutiny versus remaining open? Export control regimes compound this difficulty by imposing restrictions based on technology domain, partner nationality, and end-use in ways that create complex matrices of permitted and prohibited collaboration. After Brexit, UK universities now have flexibility to develop bilateral partnerships and position themselves as trusted partners for international research with security relevance. The goal should be enabling universities to maintain international competitiveness while managing specific security risks, recognising that technological leadership requires international collaboration and itself serves national security.

 Democratic legitimacy versus expert judgment

Public funding creates legitimate democratic claims on university decision-making about security-relevant research. Students, staff, and civil society organisations argue they should have voice in decisions about weapons development or surveillance technology. Simultaneously, research decisions require technical expertise and some insulation from short-term political pressure. The challenge is governance design: incorporating legitimate concerns while enabling institutions to fulfil societal responsibilities. The ambiguity about what constitutes defence-relevant research makes this particularly fraught. If a university community votes to prohibit weapons research, does this extend to AI algorithms that might in other circumstances have autonomous weapons applications? Or to psychology research that potentially can be used to inform military recruitment? Different communities will draw these lines differently, and governance must evaluate research based on substantive criteria rather than categorical prohibitions.

 Institutional autonomy versus funding dependency

Defence and security budgets are currently growing faster than civil research funding, creating opportunities for universities to strengthen their resource base while contributing to national security. The challenge is ensuring engagement develops through deliberate institutional choice rather than incremental drift. Investing in secure facilities, cleared personnel, and governance infrastructure creates capacity serving national interests. But it also creates path dependencies and constituencies with vested interests. The costs of exit become significant. The blurred civil-military boundary creates a particular trap, a 'boiling frog' problem. Universities might accept funding framed as civilian only to find successive iterations increasingly emphasise security applications. Each incremental shift seems manageable. Cumulatively they can transform institutional character without any single decision crossing a clear line. 

The commercialisation of dual-use technologies offers a distinct pathway. Research-intensive universities increasingly pursue spinouts and licensing arrangements that can strengthen national security by ensuring UK-developed technologies serve UK and allied interests. Areas like AI algorithms, quantum sensors, semiconductors or advanced materials all have inherently dual-use commercialisation potential. When universities license these technologies or support spinouts with security applications, they contribute directly to technological sovereignty. Defence customers often require higher technology readiness levels than civilian markets, meaning defence contractors typically act as essential intermediaries. Universities need to understand these different pathways when developing governance for commercialisation with security relevance. And the UK Government can support sustainable partnerships through multi-year funding commitments that allow universities to plan capacity development deliberately.

Transparency versus security

Academic legitimacy rests on transparency, open publication, and peer review. Some national security research requires classification, restricted access, and publication controls. These need not be fundamentally incompatible. Universities can contribute through predominantly open research while accepting limited classified work where genuinely necessary. The challenge is ensuring clarity about classification requirements before research begins, so institutions and researchers make informed choices. Classification decisions are typically made by funders rather than universities, meaning institutions can lose control over whether research remains open. Work can be retrospectively classified, creating structural constraints on academic freedom that most governance frameworks fail to address. Export controls add further complexity, particularly as their application extends to emerging technologies. Machine learning for medical imaging need not trigger controls unless specific military applications are intended. But the expanding scope of strategically relevant research means universities can find themselves unexpectedly constrained in domains they considered removed from defence concerns. Government can support this by providing early clarity about likely classification requirements and clearer guidance on which domains require enhanced protection.

The European and UK Context

 EU policy increasingly frames research through security and strategic autonomy lenses. The European Chips Act, quantum and AI flagships, and similar initiatives blend civil and security rationales. This potentially means that Horizon Europe-funded quantum research might be considered “defence research” if its explicit purpose included maintaining European technological sovereignty. Universities that would refuse national defence ministry funding may accept EU funding for functionally equivalent research because the framing differs. This suggests institutional positions sometimes rest on political symbolism rather than principled distinctions about research character. The EU’s approach exemplifies the collapse of civil-military boundaries: strategic autonomy explicitly treats civilian and security objectives as inseparable.

 The European Defence Fund explicitly funds defence and dual-use research. UK universities are excluded post-Brexit, creating both constraints and opportunities. The exclusion limits access to substantial funding but frees institutions from governance complexities that EDF participation would create. The May 2025 UK-EU Security and Defence Partnership creates a framework for closer cooperation but has not yet opened EU defence industrial programmes to UK participants. More broadly, UK universities combine established defence research traditions with flexibility from operating outside EU frameworks. This positions them well to leverage European scientific collaboration selectively while contributing fully to UK national security priorities.

 UK defence and security policy creates clear demand for research-intensive university contributions. The 2025 Strategic Defence Review identifies innovation as central to warfighting readiness, with AI, quantum, advanced materials, and synthetic biology among its priority domains. The establishment of UK Defence Innovation, consolidating the former DASA and other innovation bodies with a ring fenced £400 million annual budget, provides a single point of engagement for universities. DSTL provides research support, and the MoD Chief Scientific Adviser coordinates research priorities. Several UK institutions maintain significant national security research portfolios while others adopt more selective approaches. This diversity strengthens the national research base. Government policy can leverage it by providing clear demand signals, supporting institutional capacity development, and recognising that varied approaches create a more robust ecosystem than uniform engagement.

Implications for Governance and Policy

 Understanding these tensions enables productive partnerships. For universities, clarity enables principled decision-making about contributing to national security while preserving academic characteristics that make them valuable partners. For government, it clarifies what policy support universities need and where barriers to engagement can be removed.

The two fundamental ambiguities mean categorical approaches are no longer sufficient. Practical governance frameworks might include several elements. First, presumptive positions with case-by-case override. Presumptions should be established with clear criteria and processes for exceptions, acknowledging that categorical rules cannot address inherent ambiguity. Second, transparency about uncertainty: acknowledge publicly when partnerships fall in genuinely ambiguous territory, explaining competing considerations and institutional reasoning. This builds legitimacy even when stakeholders disagree. Third, dynamic review mechanisms: that accept strategic relationships shift and dual-use applications emerge, building periodic review into major partnerships. Fourth, distributed expertise: that develops institutional capacity for strategic assessment beyond traditional research ethics review, including expertise on geopolitical risk, export controls, and technology security.

Government itself faces analogous tensions. Defence spending competes with other priorities. In devolved nations, the Scottish Government's position on nuclear weapons creates tensions with UK defence policy even as nuclear submarines operate from Scottish bases. Recognising that government navigates its own structural tensions, rather than speaking with a single voice, should inform how universities and government engage on these issues. And it is well established that multiple government departments have stakes in university research, sometimes with competing or even conflicting objectives. Productive partnerships require acknowledging complexity on both sides.

 For Parliament and the government, several implications follow:

• Expect and accommodate diversity in university responses. Research-intensive universities will respond differently to defence research opportunities based on their traditions, missions, and governance structures. Policy should treat variation as ecosystem strength rather than institutional obstruction.

• Understand structural constraints. Universities cannot easily segregate “defence” from “civilian” research or “friendly” from “adversarial” partnerships when fundamental categories have eroded. Policy demanding such segregation will produce arbitrary distinctions disconnected from actual security implications.

• Recognise inherent trade-offs. Deeper defence partnerships may constrain international collaboration, research security restrictions may reduce access to international talent, and transparency requirements may conflict with classification needs. The temptation to demand both deep engagement and unrestricted openness sets universities impossible tasks.

• Support institutional capacity development. Universities need expertise to navigate geopolitical risk, export controls, and strategic assessment. Most institutions currently lack resources to conduct sophisticated analysis of which partnerships pose genuine security risks versus which restrictions are administratively convenient but substantively arbitrary. The Trusted Research agenda and Research Collaboration Advice Team represent progress, but sustained investment in such mechanisms remains essential. Government should also consider how to support universities in building the rapidly evolving compliance and assurance infrastructure that sophisticated security assessment demands. This requires resources, not merely policy direction.

• Finally, create space for ongoing dialogue as the landscape shifts. New technologies emerge, strategic relationships change, and security threats evolve. Regular structured engagement helps both sides understand evolving constraints rather than allowing policy and practice to drift apart until crisis forces confrontation.

UK research-intensive universities possess world-leading capabilities in domains critical to national security. Enabling their contribution requires clear demand signals, investment in institutional capacity, multi-year funding commitments, and recognition that varied institutional approaches create robust national capability. The path forward requires intellectual honesty about the complexity of contemporary security research, and mutual recognition that universities and government share interests in strengthening UK security and technological advantage.