Aspen Digital

Executive Summary

In October 2024, Chatham House and Aspen Digital convened events in London and Washington, DC, bringing together a select group of over 50 high-level US and UK policy decision-makers, shapers, and influencers. These events were designed to better understand what we mean by competitiveness in the context of an “AI race” with China and the implications not only for the US and UK, but also for other parts of the world.

Much of the discussion around AI competition with China in the US and UK is focused on national security, trade, and technical research and development (R&D). To interrogate this framing and move beyond the usual narratives about competing with China, participants were encouraged to adopt a long-term, strategic approach to AI competitiveness, focusing on what a “grand strategy” in the age of AI might look like. This more expansive frame considers the long-term priorities of a country, which includes military and national security interests, but also emphasizes prosperity and societal priorities. 

In exploring and adapting this broader grand strategy frame, participants engaged with a wider range of inputs to competitiveness and developed more expansive and realistic understandings of what “winning” and “competing” mean in different national contexts. Participants in both meetings highlighted how the strategic application of domestic policy levers (e.g., talent and workforce policy), in addition to traditional security and trade approaches, can improve US and UK AI competition with China. They also highlighted the need for greater alignment on responsible and trustworthy AI governance standards.

While the US and UK are allies and share some priorities and values as democratic, diverse societies, there are significant differences in their approaches to competitiveness. Often considered a middle power, the UK is by no means competitive with the US or China when it comes to AI development, but uniquely excels in certain domains, like convening authority and talent. The US, on the other hand, boasts an often unparalleled technological and economic competitive edge in AI, but is ceding important international allies to China as a result of its lax stance on tech industry regulation and aggressive trade barrier enforcement.

The view from London

8 October 2024

The first roundtable explored the UK and China’s competitiveness on AI, examining what the drive behind competition reveals about the UK’s geopolitical aspirations and positioning, while also addressing its domestic policy priorities.

Held three months after the UK general election, the event welcomed a diverse group of opinion-shapers and experts from: policy and government; security, intelligence, and defence; research and academia; civil society and advocacy; and the private sector. The convening was held under the Chatham House Rule.

With welcoming remarks, two plenary sessions, and issue-focused breakout sessions (on the technology, trade, and talent “drivers” of competitiveness), Rethinking UK-China Competitiveness on AI aimed to establish a shared understanding of what makes the UK—and other diverse democracies—competitive. The following event summary captures areas of agreement and divergence in the characterization of the UK’s net competitiveness vis-à-vis China, in addition to platforming several takeaways specific to the UK.

Understanding competitiveness

The event mapped out a diversity of perspectives on the UK’s strategic approach to competing with China on AI, but also on the UK’s competitiveness overall. Participants were encouraged to adopt a bird’s-eye view and consider how UK competitiveness might be judged from Beijing. 

AI Competition with China

The UK’s strategic approach to China was front-and-center of the discussion. From a government perspective, the approach aims to be clear-eyed, consistent, long-term, and rooted in UK priorities and global interests. This dovetails with Labour’s Manifesto and the UK’s government’s proposed audit of bilateral relations and its so-called “three C’s approach” to engaging with China: cooperate, challenge, and compete. This approach has gathered considerable criticism rooted in frustrations about the lack of transparency, information-sharing, and clear directionality within the UK’s China policy.

Some participants called for implementing a more pragmatic assessment of the UK’s “endgame” when it comes to competing with China. In no uncertain terms, the UK cannot outcompete China when it comes to developing emerging technologies like AI. The UK must be realistic about its lack of relative competitiveness and the presence of existing benefits derived from cooperation on technology. It is essential to avoid flattening the narrative about UK-China competitiveness into a “technology arms race.” Participants from government and defense also invited practical, evidence-based challenges to the government’s policy on China and AI. It was suggested to focus on seeking a “common cause” with China in pushing the needle on AI safety, in particular, and identifying concrete policy options for how the UK can navigate the “compete vs. cooperate” dilemma.

Reframe the narrative

The UK’s global role is changing. As a middle power with diminished global influence, it is increasingly constrained by a web of interdependencies that it relies on for the development of emerging technologies, including both China and the US. While the UK fails to outpace China on AI (for example, on computation and data), the country nonetheless wields immense convening power as a global “moderator” and attracts (although perhaps struggles to retain) global talent, predominantly through influential higher education institutions.

To this end, understanding UK AI competition with China requires a two-step shift. First, it demands honesty and pragmatism about UK technological and market capabilities and geopolitical agency. Second, it demands reframing the competitiveness narrative to consider domestic needs and priorities (“roots”). How does this bear on UK expectations to build global capabilities as a “science and technology superpower,” as outlined in the 2023 International Technology Strategy?

As an example, the UK’s industrial strategy lacks China’s long-term “strategic patience.” The UK cannot outspend China on, say, R&D, but it can improve targeted investments in its domestic AI investment ecosystem. Doing so requires building better industrial strategy and spending decisions that are responsive to domestic constraints, opportunities, and aspirations.

How to strengthen domestic roots

There are pervasive uncertainties about what “UK values” are, how they apply to technology, and specifically the use of technology for public ends. Despite this, the event’s conversation circled around a common question: what is the ideal vision for a strong, credible, and democratic state when it comes to the use and deployment of technology, and how does this bear on competitiveness?

A shared takeaway was the importance of adaptable institutions and innovation in governance as part of the bedrock for both improved state capacity and improved competitiveness on AI. While AI is by no means a “silver bullet,” participants noted it has the power to capture public imagination and provide an urgency for (re)building more agile institutions. As one participant put it, “it’s not about racing, but rather ‘turning inward’” to consider the societal challenges improved public capacity in AI can help tackle. Momentum for advancing public capacity in AI can also be derived from the perceived presence of security and safety risks, although over-securitization may prove damaging to public trust. 

Driving competitiveness: a path forward for the UK

The event included breakout discussions on drivers of competitiveness in three interlinked issue areas: technology, trade, and talent. These sessions’ respective starting points were small-scale case studies on semiconductor chips, export controls, and security concerns around reliance on global talent. They revealed paths forward for the UK to marry “turning inward” with deriving benefits from globally-focused technology ecosystems, talent, and trade, while also drawing on lessons from China’s experience in each area.

Bolster tech ecosystems

As a central component of AI competitiveness, the UK aims to build a homegrown chip industry and embed itself into global semiconductor supply chains. But compared to US, EU, and China’s spending on chips, the UK’s currently falls short. The UK is unlikely to achieve fully sovereign chip capabilities. Recognizing this, one participant noted the importance of building “friendships.” By integrating into a globally diverse semiconductor ecosystem—referred to as a “web of middle powers”—the UK can enhance stability and sustained access to technology.

The case of Cambridge’s ARM computing technology might exemplify how the UK can draw on its domestic strengths and global networks to build competitiveness; the effort was initially and remains an industry-academic collaboration, inclusive of different countries and with UK values streamlined throughout. Participants also called for embedding UK companies and universities into the chip lifecycle through targeted interventions in IP and technology design support.

Retain diverse talent 

UK national conversations increasingly center emerging concerns about talent from abroad and the potential security risks this introduces for developing AI. These concerns fall along a spectrum ranging from pragmatic, evidence-based arguments to alarmist rhetoric. On the flip side, the UK is home to considerable consensus about the power of diverse, global talent for building a thriving AI ecosystem and improving competitiveness. Participants agreed the UK is adept at creating a competitive ecosystem (through its capital markets, sandboxing, and talent pools, for example) but falls short in retaining talent.

Participants additionally discussed the need for a values-driven visa policy for attracting and retaining globally competitive talent and their outputs, such as university spinouts. Investing in universities and strengthening the UK’s identity as a hub for academic-industry engagements on emerging technologies must complement these efforts.

Here, lessons can be drawn from China, a country increasingly retaining homegrown AI talent while also developing knowledge-exchange partnerships and investment for attracting talent from the global majority.

Clarify exports and trade policy

The UK confronts the unique challenges of a post-Brexit trade environment, where the complex rules underlying trade policy are being challenged and rewritten. The new government may also change the UK’s position. For instance, the Starmer government’s audit of UK-China relations will have implications for how the two countries engage on trade, and particularly trade on dual-use technologies. Participants agreed that export controls on AI-relevant components have had mixed success.

Trade policy is a lever for attaining the UK’s long-term ambition of developing strategic autonomy, which participants judged to be a key part of competitiveness. But operating this lever demands “turning inward” and tackling complicated questions, such as: where should the UK draw the line when it comes to controlling the security risks associated with foreign-owned technology companies and their products? The Huawei 5G incident (2019) was cited, underlying the UK’s lack of maneuverability: a lack both manufactured by policy shortcoming and imposed by geopolitical constraints.

Build an evidence base for improved policy 

While more transparency from policymakers on inputs in China and AI policy is required, researchers also have a critical role in producing and disseminating opinion-shaping material to policymakers. Part of this task is seeking to expose and inform how influential actors define geopolitical power as emerging from AI, as these understandings can concretely impact industrial policy: for example, whether AI’s impact on society is understood as disruptive and distinctly technological, or diffusive. As discussed above, there are a variety of competition-relevant policies that fulfill an inward-facing agenda, responsive to domestic priorities and aspirations.

Reflections

In early 2025, the UK Government will announce its findings from its audit of UK-China relations. The roundtable underlined the urgent need for building a more pragmatic, realistic approach to bolstering UK competitiveness, both domestically, vis-à-vis allies, and vis-à-vis competitors, like China. While the audit itself has attracted criticism (as outlined above), the announcement of its findings will be an opportunity for UK decision-makers—on emerging technologies and on China policy—to strategically reorient the national “conversation” to focus on the public policy levers underlying competitiveness, and the national aspirations these can help achieve.

The view from Washington 

29 October 2024

Like the London convening, the second roundtable in Washington, DC convened experts across sectors to discuss US strategic objectives in AI competition with China. Attendees were asked to consider the underlying dynamics of this competition, including its geopolitical, economic, and societal dimensions, and to reflect on what “winning” would entail and yield.

Participants were encouraged to reframe competition through the lens of a “grand strategy” informed by American democratic values. Participants also discussed how the US’s competition-focused trade policies have second order effects that implicate traditional allies such as the UK, as well as nations in the global majority. The discussion underscored the need for a nuanced approach to competing with China that balances national security considerations with global cooperation, as well as the need to strengthen domestic resilience to ensure that Americans are helped, not hurt, by the big societal changes that widespread AI deployment will bring in the coming decade.

Reframing the narrative through the lens of grand strategy

In the US, many of the threads of the US-China AI competition debate are tangled around national security concerns about expanding AI capabilities. These concerns are reflected in policies on military use of AI, public investment in R&D and domestic semiconductor manufacturing, aggressive trade policies, and efforts to protect US intellectual property. 

Proponents of the national security framing of competition advocate a first-mover mentality, arguing that “winning” requires building the biggest and most-capable AI models as a form of deterrence. Unfortunately, this winner-take-all approach can preclude international cooperation and ultimately pose greater societal risks. For this reason, some participants with expertise in nuclear disarmament raised concerns that this securitized frame for AI competitiveness could hinder important international collaboration on AI governance, in the same way that military competition has impeded nuclear safety goals. 

This is not to say that national security concerns are irrelevant. However, there is already a robust national security policy conversation taking place regarding competition on AI. In this event we chose to put that aside to interrogate what opportunities for US competitiveness we might be missing or might be crowded out due to an overreliance on the securitization frame. (Since the roundtable, the introduction of the open-weight models from Chinese company Deepseek have upended many of the core assumptions about what competition “must” look like.) By using the concept of grand strategy as a reframing tool, participants considered what domestic policies the US could pursue that would support the nation’s competitiveness in areas besides raw AI capability.

Participants dug into reframing competitiveness using the following key democratic values and discussed how to realize them through policy action: equity, opportunity, liberty, and global cooperation. 

For each value, the following AI competition strategies were identified:

1. Equity: encourage adoption of AI by ensuring that the future of AI includes everyone;
2. Opportunity: make it possible for more Americans to innovate with and benefit from AI;
3. Liberty: safeguard individual rights, promote innovation within competitive markets, and strengthen our democracy through government modernization; and 
4. Global Cooperation: counter authoritarian narratives by spreading democratic values globally and building and benefiting from existing collaborative international frameworks.

Equity: build the inclusive foundations of AI adoption

Fear of labor displacement is a significant risk to AI adoption and the productivity benefits it could bring. The mounting public rejection of AI could undermine competitiveness by limiting the “diffusion,” or uptake, of this strategic technology. To mitigate the workforce disruptions caused by AI, participants emphasized the need for equity-driven domestic policy measures that build robust safety nets and foster workforce adaptability. 

Expanding retraining programs through public-private partnerships was identified as essential to transitioning workers into AI-related roles, such as maintaining advanced machinery. Equally important is investing in STEM education and diverse learning opportunities to ensure inclusivity in AI-related fields. Additionally, providing grants for small business use of AI tools and vocational training programs was recommended to bridge skills gaps and promote economic empowerment.

Together, these initiatives aim to distribute AI’s benefits equitably, making Americans more willing to embrace these technologies, which enhances the nation’s ability to compete with China and other global powers.

Opportunity: unlock innovation potential for all Americans

One of the better-recognized facets of US-China competition is the economic competition that has primarily manifested as trade negotiations and export controls. However, core to economic competitiveness is the strength of the innovation ecosystem at home. This requires focus on both infrastructure and talent.

Participants underscored the transformative potential of AI for small businesses and startups, advocating for a supportive environment where grassroots innovation can flourish. These smaller players need infrastructure that can support their growth. Establishing and promoting standards for data sharing and interoperability (such as the National Secure Data Service Demonstration) and expanding public AI initiatives (like the National AI Research Resource Pilot or state initiatives like Empire AI) are ingredients of technological competitiveness and can broaden access for smaller players. Open source AI development was also recognized as a key form of infrastructure that should be supported to promote grassroots innovation.

Leaning into the benefits of colocated resources and talent in innovation hubs like Silicon Valley was also proposed. Today, many of these places have become so unaffordable that they are losing the synergies that made them effective in the past. Investments in affordable housing and infrastructure within tech hubs were identified as crucial to attracting and retaining diverse talent in these high-yield ecosystems that are central to economic competitiveness. Strengthening the global talent pipeline through immigration policies, such as an expanded “genius visa” program, was also mentioned as critical for promoting competitive innovation at home.

Liberty: safeguard privacy and promote fair competition

In order to secure the benefits of equity and opportunity as discussed above, participants recognized that the US must also prioritize advancing liberty so as to not undermine gains in competitiveness.

For example, while data interoperability was emphasized to promote innovation, participants also noted how uncritically promoting data access can provide advantages to geostrategic competitors by accelerating their AI development. Privacy intrusions and the threat of manipulation (through misleading AI-generated content) also undermine the public’s trust in AI tools, increasing public resistance. Some participants recommended enacting federal privacy standards inspired by General Data Protection Regulation (GDPR) to safeguard American data from foreign adversaries and promote civil liberties that cultivate public trust.

Similarly, while investing in grassroots innovation is important, fair market standards are crucial to ensuring that smaller players have an opportunity to compete. The AI industry has demonstrated a tendency to concentrate, largely due to the immense costs of compute, data, and talent, which create high barriers to entry. Addressing these challenges is essential to fostering a more diverse technology ecosystem that can safeguard against negative disruption while ensuring broader access to AI innovation. It was also noted that market concentration presents a national security risk, as individual points of failure could produce system-wide consequences.

Participants emphasized that in order to safeguard privacy and promote fair competition we need effective government in the United States. As such, it is critical to take actions to make government responsive in the face of technological change, such as supporting programs that modernize Congress’s digital infrastructure.

Global cooperation: build strategic partnerships

Participants agreed that the US must not approach AI competition with China in isolation, but rather should cultivate a robust and sustainable network of partners. They stressed the importance of engaging democratic allies and nations in the global majority by fostering partnerships on equal footing that reflect shared interests. Trust-building requires aligning promises with tangible benefits, such as supporting local innovation hubs and addressing local infrastructure and economic priorities. Multilateral forums like the G7 and OECD are critical venues for affirming a shared framework for cooperation, addressing data privacy, and advancing other collaborative initiatives.

Recognizing that many nations seek strategic autonomy, the US must advance partnerships that respect sovereignty while offering attractive alternatives to authoritarian models of technology. This is especially important in the face of low- or no-cost alternatives like China’s DeepSeek. Making American AI readily available to the world through open weight and open source models is an important step toward advancing democratic values.

Reflections

Roundtable participants embraced a strategic, values-driven framework that secures the US and its allies’ competitive edge while promoting long-term global stability. A central component of successful competition is economic strength. For the US to build a robust foundation for long-term economic competitiveness with China, it must enact initiatives that build public trust in emerging technologies like AI and enable widespread adoption. By leveraging democratic principles, fortifying domestic foundations, and fostering global collaboration, the US can not only outcompete but outlast authoritarian alternatives. This approach ensures that AI leadership aligns with humanity’s broader interests, reinforcing a vision of technology as a force for inclusive and ethical progress. 

This vision is especially crucial given the results of the US elections and the new administration’s approach to international relations. In the face of increasing polarization, reinforcing democratic values and international cooperation becomes even more critical to shaping AI futures.

Transatlantic steps

While there were several overlapping strategic priorities for transatlantic partners, the following are particularly promising for future dialogue and convening.

Foster and protect strategic and scientific dialogue with China

The severing of institutionalized communication among the US, UK, and China has created uncertainties, tension, and gaps in mutual understanding. A clearer understanding of national positions and priorities can strengthen diplomatic channels and reinforce the democratic principles of transparency and accountability in emerging technology issues. On AI safety especially, the US and UK would benefit from coordinating approaches to China, given both countries have demonstrated success in engaging China in bilateral and multilateral arrangements, such as the Bletchley Park Summit and US-China bilaterals in 2024.

All three countries have a track record of productive knowledge exchanges and dialogues on AI. The UK government in particular views science and technology as “enablers” within many areas of potential cooperation with China. Scientist-led or -driven exchanges are an opportunity for strategic dialogue with China due to their depoliticizing potential and power of interpersonal connections; in fact, a science-centric approach to competitiveness would platform the significance of the scientific method in encouraging healthy competition on the basis of information exchange and equitable access to scientific data for progress.

Engage and strengthen democratic alliances in the global community

The US and the UK both engage middle and emerging powers including global majority countries in standard-setting and AI governance. It is important that the US and UK continue to foster and deepen relationships with these other nations and not cede diplomatic relations to China. Both the US and UK must chart a path between building values-driven alliances while also strategically engaging with China on areas of urgent and shared interest, such as AI safety. 

Existing international arrangements for global AI governance—from international standards bodies, research partnerships, and summits—should be leveraged to forge deeper, values-based partnerships with other democracies. They should also be used to engage with emerging and middle powers from the global majority on equitable, inclusive footing. Within these venues, democratizing inputs into global AI governance should be considered a vehicle for both strengthening multi-stakeholderism as a principle and as a counter-weight to authoritarian models of technology governance.

Acknowledgements

This work was made possible with the support of the Center for Responsible Innovation, and was produced by Andrea Chiampan, Morgan Engel, and B Cavello from Aspen Digital and by our collaborators and coauthors from Chatham House’s Digital Society Programme: Alex Krasodomski-Jones, Isabella Wilkinson, and Rowan Wilkinson. We’d also like to thank Vivian Schiller, Isabella Sarmiento, Beth Semel, Devon Regal, and Eleanor Tursman for their support.

Share your thoughts

If this work is helpful to you, please let us know! We actively solicit feedback on our work so that we can make it more useful to people like you.

Strategic Reorientation on AI Competition with China © 2025 by Aspen Digital is licensed under CC BY 4.0.