Published on 24 April 2024
Share Tweet  Share

How geopolitical tensions are shaping the future of AI

Geopolitical tensions are on the increase and the governance of AI is at a crossroads. In one direction is a future shaped by a narrative about an AI arms race between warring powers. In another, the technology has the potential to develop in a way that benefits societies across the globe. As AI’s transformative potential becomes clear, the new Bennett Institute project, AIxGEO, will analyse the shifting narratives regarding AI and map the potential for a collaborative future. Aleks Turobov explains.

In today’s context of geopolitical uncertainty, the narratives surrounding Artificial Intelligence (AI) have increasingly shifted towards its potential for national gain and militarisation. The policy debate often focuses on security issues and the spectre of a so-called ‘AI arms race’, invoking images of technologically enabled dominance in defence, the economy and even in political control. Meanwhile, these fears are prompting political interest in more intervention in the regulation of AI.

Beneath these debates lies a deeper concern about AI’s transformative power across the social, economic, and political spheres. States want to control technology and guide innovation for their national interests and at the same time do not want to be left behind in its development.  Notions of technological sovereignty dominate policy as a result, seen not only in the dominant AI powers–the United States (US) and China–but even in the European Union’s (EU) focus on digital sovereignty and security integration. This focus on sovereignty risks suppressing some innovations or constraining the technology’s broader benefits for society.

It is possible, however, to imagine seeing AI governance through a different lens, putting economic and societal needs at the forefront of the debate or centring human rights. This may help us break free from the current cycle of AI related fears and restrictions, moving instead towards a collaborative regulatory environment that respects and recognises multi-stakeholder interests across different countries, paving the way for an AI future that is not only resilient but also inclusive of all.

AI and power dynamics

The narrative of AI as a matter of international rivalry emerged from a growing sense that this particular technology would have an outsized impact on the world’s future balance of power, particularly after DeepMind’s seminal AlphaGo moment and Xi Jinping’s announcement that China would be an AI superpower by 2030, which some viewed as an indirect response. Gradually the movement earned followers and acolytes, morphing into a battle for semiconductor supremacy and into the realms of cloud computing. Since the release of ChatGPT and resultant hype around what are often termed ‘foundation’ or ‘frontier’ models, the acceptance of AI as the latest frontier in the geopolitical contest has only seemed to cement. During this rivalry, semiconductors have emerged as critical to enhancing computing capabilities, essential for advancing major digital technologies, including 5G, 6G, cloud computing, and robotics, further intensifying the global balance of power.

Chips came further into the spotlight at the height of the global supply chain disruption during the COVID-19 pandemic. The US semiconductor industry contributes 39% of the total value of the entire global semiconductor supply chain from chip design to shipping the final product, and its allies, Japan, Netherlands, United Kingdom, Germany, Taiwan and South Korea, collectively contribute around 53%. China, in stark contrast, contributes just 6% (Khan et al., 2021). However, Taiwan dominates advanced chip fabrication and the Netherlands the advanced photolithography machines so there are clear vulnerabilities. This makes these supply chain elements particularly sensitive to geopolitical tensions and disruptions from concentration in a few markets. Such geopolitical tensions are leading both the US and China to invest in enhanced domestic semiconductor manufacturing facilities.

This is a costly feedback loop: supply chain disruptions encourage the political and economic process of de-globalisation and ‘sovereignisation’, further disrupting international chip logistics and global markets. The EU and US Chip Acts demonstrate the strengthening of such trends, and a shift away from market-oriented liberal policies towards intervention-oriented techno-nationalism, with a clear prioritisation of geopolitical motives (Luo and Van Assche, 2023) in national policy and strategy. This narrative around AI is framed in terms of national security and superiority, thereby shaping the policy landscape and reproducing an unidimensional understanding of AI that can lead to damaging consequences.

This is why it is time to explore and unpack the current debates and to interrogate and challenge accepted truths. It’s important to provide alternative visions and policy recommendations that aim to foster global collaboration, address the multifaceted impacts of AI, and shift the policy debate towards a more cooperative future of AI development. This is the focus of the new AIxGEO project, established by Director Verity Harding at the Bennett Institute for Public Policy, which I have recently joined to work full time on such issues.

Achieving a counter-narrative to the ‘AI arms race’?

As I begin working on AIxGEO, my early research will focus on examining the policies of major international organisations like the UN (United Nations), the OECD (Organisation for Economic Co-operation and Development), NATO (North Atlantic Treaty Organization), the WTO (World Trade Organization), and the EU. I aim to understand the key themes, concerns, and challenges that are shaping the global AI discussion by conducting a thematic analysis of the many hundreds of policy documents.

Next, I will connect the dots between these organisations by identifying common themes and shared narratives about AI, using network analysis. This analysis will spotlight existing and potential coalitions that are united around specific AI viewpoints. Such insights will pinpoint key stakeholders and their concerns, facilitating a collaborative approach to AI governance. A better understanding of specific narratives and international support provides practical common ground and entry points for multi-stakeholder engagement.

AIxGEO project

The ultimate aim of the project is to provide foundations for stakeholders to cooperate in shaping the development of AI. As we stand at a crossroads in AI advancement and geopolitical dynamics, the AIxGEO initiative will help navigate the multifaceted challenges in AI policy through a pragmatic and evidence-driven approach. Putting human rights, democratic principles, economic outcomes and multi-stakeholder involvement at the heart of AI policy will highlight the opportunities to deploy AI as a technological tool for societal benefit.

The views and opinions expressed in this post are those of the author(s) and not necessarily those of the Bennett Institute for Public Policy.


Dr Aleksei Turobov

Research Associate

Dr Aleksei Turobov is a Research Associate working on the AIxGeo project at the Bennett Institute for Public Policy. His research centres around the nexus of AI policy, politics, and...

Back to Top