Written evidence submitted by Bennett Institute for Public Policy (SDY0026)

The AI & Geopolitics Project (AIxGEO) at the Bennett Institute for Public Policy, University of Cambridge, is a research initiative dedicated to analysing the intersection of Artificial Intelligence (AI) governance and international relations. Our evidence-based research offers valuable insights into how the UK can strengthen its position as a global science leader while navigating complex geopolitical challenges.

Our report, "Moving beyond competition: domain-specific approach for international AI framework," represents the first comprehensive analysis of AI governance approaches across major international organisations; we have systematically mapped the evolving landscape of international AI governance and identified concrete opportunities for productive cooperation, even amid strategic competition.

The UK's distinctive position in the global technology landscape creates a unique opportunity to shape international norms and standards while advancing national prosperity and security. Our research directly addresses the Committee's interest in understanding how science and technology can enhance the UK's international influence, contribute to national security, and support economic growth through global R&D investment. This submission addresses the Committee’s questions for which our work provides evidence.

To what extent is the UK considered a global leader in science and innovation and how does this contribute to its soft power on the global stage?

The UK is uniquely positioned in the global AI landscape, combining world-leading research capabilities with a balanced regulatory approach. This distinctive position creates an opportunity for the UK to serve as a mediating force between different global regulatory frameworks - particularly between the US's more market-oriented approach and the EU's more comprehensive regulatory paradigm as embodied in the EU AI Act. Of particular interest to our research at AIxGEO is how this positioning enables the UK to play a strategic mediating role on the global stage in what has become an increasingly consequential technology geopolitically, potentially bridging divides between competing AI governance visions while advancing its innovation interests and values-based approach to AI development.

Our research indicates that nations that combine technical expertise, balanced governance approaches, and diplomatic networks can serve as 'honest brokers' in the international AI ecosystem. With its strong tradition of scientific excellence and diplomatic engagement, the UK is ideally positioned to play this role. As early as 2017 the UK has offered important contributions to key international AI governance initiatives, such as the first set of OECD AI Principles in 2019 and the G7 process which established the Global Partnership on AI in 2018. This has enhanced the UK's soft power by establishing the country as a thought leader and pragmatic voice in global AI governance discussions. Combining scientific expertise with diplomatic capacity and developing best practices in AI-driven civil service reform, the UK's influence in AI policy discussions can improve the national brand and geopolitical position.

How has this agenda been impacted by the current geopolitical environment, including the international activities of Russia and China?

The current geopolitical environment has significantly influenced AI governance discourse, often framing it as a zero-sum competition between major powers. Our research identifies a problematic narrative of an "AI arms race," particularly between the US and China, which threatens to fragment the international order and diminish opportunities for beneficial collaboration, while minimising both the perception and the reality of the influence of other nations, including the UK, in the debate.

The evolution of AI policy has increasingly been shaped by security rationale and the perceived threat from 'foreign entities of concern.' Major global players like the US, China, and the EU are focused on building national champions and ensuring technological sovereignty, particularly in critical areas like semiconductors and AI foundation models. For the UK, uncritically adopting this zero-sum competitive framing risks undermining its distinctive diplomatic position and constraining its ability to influence global AI governance trajectories. By reflexively aligning with security-first narratives, the UK would sacrifice its potential role as a credible bridge-builder capable of advancing cooperation in domains where mutual benefit outweighs competition. The evidence indicates that nations maintaining flexibility in their diplomatic positioning ultimately secure greater influence in shaping emerging technologies' rules and norms than those who position themselves primarily as competitors.

This geopolitical tension has also led to a proliferation of export controls, investment screening mechanisms, and other trade restrictions that risk creating technological 'blocs' and undermining global scientific collaboration. The semiconductor industry, critical to advanced AI development, has become particularly politicised, with measures such as the US CHIPS Act and export controls on advanced chips reflecting growing techno-nationalism.

Despite these challenges, our research identifies opportunities for constructive engagement even in this tense geopolitical environment. The domain-specific approach we advocate recognises that collaboration in areas such as environmental monitoring, healthcare applications, educational technology and science diplomacy can proceed even as strategic competition continues in other domains. This nuanced approach allows the UK to maintain security vigilance while advancing beneficial international cooperation.

How does science and technology innovation contribute to the UK's national security?

Science and technology innovation contributes to the UK's national security in multifaceted ways, extending beyond traditional defence applications to include economic security, critical infrastructure resilience, and the ability to shape international norms and standards.

Our research identifies two key pathways through which AI and related technologies enhance national security:

Risk mitigation through established frameworks. The UK's AI safety and ethics approach provides frameworks for systematically addressing security concerns. Rather than restricting innovation, domain-specific and balanced (state control over data and critical infrastructure with supply chain diversification with like-minded partners) governance approaches can ensure security is built into technological development.
Maintaining technological competitiveness while enabling collaboration. By focusing on specific domains where cooperation delivers clear benefits (such as cybersecurity, environmental monitoring or agriculture), the UK can maintain technological advantage in critical areas while benefiting from international collaboration.

In cybersecurity, our research shows a particular opportunity where shared vulnerabilities create compelling imperatives for international cooperation. NATO's landmark determination that a cyber-attack can trigger Article 5 marks a fundamental shift in security doctrine. In contrast, the growing recognition that digital security risk undermines trust in digital transformation creates a strong rationale for international cooperation, even among strategic competitors. The UK's leadership in AI safety research positions it to play a pivotal role in developing international norms around secure AI development. By implementing its advanced capabilities in AI safety research, the UK enhances its national security while contributing to global security frameworks.

How effective is the UK Government's strategy for positioning the UK as a global leader in science and technology and what role does the Department for Science, Innovation and Technology (DSIT) play in advancing this agenda?

There are several areas where the DSIT strategy could be strengthened - most notably in refusing to accept the prevailing paradigm that positions AI as an arms race. Doing so only hampers the UK’s ability to influence and shape events. Instead, DSIT should champion a domain-specific approach to international cooperation. Rather than pursuing unrestricted openness or comprehensive restrictions, the UK should identify specific domains where international collaboration benefits all parties. The UK can partly achieve this by:

Looking beyond Silicon Valley. Engagement and relationships with the US scientific community are important, but they should not come at the expense of investing in other global scientific allies - most notably the EU and India, as well as technologically advanced regional leaders such as Singapore, South Korea, Japan, etc.
Leveraging existing frameworks. Our research demonstrates how existing international organisations (UN, OECD, WTO, NATO) provide complementary frameworks that the UK can leverage for international engagement.
Developing robust measurement frameworks. The impact of science diplomacy activities should be systematically evaluated using metrics that capture both short-term outputs and long-term relationship building.

Does the UK need an International Science Strategy and what would it contain?

We agree that the UK would benefit from a comprehensive International Science Strategy that moves beyond the current framing of technological competition to embrace a nuanced approach.

Such a strategy should include:

A domain-specific cooperation framework that identifies areas where collaboration delivers clear benefits while minimising security concerns:

● Agricultural development: addressing food security and sustainability

● Educational technology: enhancing learning while respecting sovereignty

● Cybersecurity: converting shared vulnerabilities into collaboration frameworks

● Environmental monitoring: leveraging shared climate challenges

● Healthcare innovation: combining technical advancement with universal access

Strategic complementarity approach that recognises how distinct organisational capabilities can be combined to enhance overall policy effectiveness. Rather than duplicating efforts across government departments, the strategy should leverage the unique strengths of different actors in the UK system.
Distributed resilience model that maintains multiple centres of expertise and responsibility - technological challenges of increasing complexity are best addressed through networks of specialised capabilities rather than centralised control.
Clear principles for balancing openness and security considerations. The strategy should provide guidance on when to prioritise international collaboration and when national security interests require more restrictive approaches.
Institutional mechanisms for coordination across government departments, research institutions, industry, and international partners. Effective science diplomacy requires the alignment of diverse stakeholders around common objectives.

How well positioned is the government to link scientific and technological progress with enhanced global and UK security?

The UK's balanced approach to AI governance positions it to bridge between different regulatory approaches. British leadership in AI safety research provides substantial credibility in international discussions on responsible and trustworthy AI development. Furthermore, moving toward domain-specific AI policy frameworks, rather than pursuing general and abstract overall AI regulation, will demonstrate a practical commitment to nuanced governance that can accommodate innovation and security concerns while facilitating international cooperation in less sensitive domains.

Despite these advantages, several challenges exist in linking scientific progress with security outcomes. The prevailing narrative of technological competition increasingly overshadows valuable opportunities for security-enhancing cooperation, potentially limiting diplomatic flexibility. Additionally, persistent institutional silos between scientific and security communities impede the coordination necessary for coherent policy development, with different vocabularies, priorities, and risk assessments creating barriers to collaborative approaches. Perhaps most concerning is that measurement frameworks for evaluating the security impacts of scientific cooperation remain underdeveloped, making it difficult to demonstrate the security benefits of international collaboration empirically.

Adopting a "cooperative advantage" framework that recognises security benefits can flow from carefully designed international collaboration (starting with minilateral agreements) rather than seeing security as requiring isolation. At the same time, building institutional bridges between the scientific community and security establishments to foster mutual understanding and coordinated approaches that maintain the UK's competitive edge while maximising cooperative opportunities in appropriate domains.

9th May 2025