Bharath Reddy and Pranay Kotasthane

Abstract

Artificial intelligence has become an area of geopolitical contestation, and the policy

instruments states reach for are shaped by the narratives through which the technology is understood. This chapter identifies five competing framings – AI as a conventional arms race,

as a nuclear weapon, as a geopolitical innovation race, as a general-purpose technology, and

as a global commons. It shows that the narrative that dominates a given conversation often

reveals political and strategic objectives rather than mere technical realities. It then maps these

framings onto the instruments deployed across the AI stack: export controls and monitoring

mandates at the compute layer, localisation requirements at the data layer, and weight

restrictions and openness debates at the model layer. It concludes with pathways for middle

powers like India to mitigate near-term risks while building long-term innovation capacity.

Introduction

This essay argues that the technological and policy ecosystem surrounding artificial intelligence (AI) has become a central arena of geopolitical contestation. It begins by exploring the powerful and competing narratives that shape national AI strategies and policy choices. Whether framed as an arms race, a nuclear weapon, a geopolitical innovation race, a general-purpose technology, or a global commons, each narrative surrounding AI advances particular interests while constraining others. The dominant narrative in any given policy debate often reveals more about the geopolitical stakes of AI than the underlying technical realities it purports to describe.

Technology has long been a key determinant of economic and military power, and narratives that emphasise the links between technology and national power have become increasingly influential. Nuclear weapons reduce the likelihood of large-scale conventional wars, and globalisation has made economic decoupling prohibitively costly, if not entirely unfeasible. Although recent developments challenge these assumptions, the past few decades have seen relatively free trade and limited conflict. In this landscape, technology is becoming both a means and an object of geopolitical contestation. 

Artificial intelligence is reshaping both the sources and applications of national power – both economic and military. It introduces new elements of power, changes the importance of existing sources of power, and alters the goals that states should pursue. For instance, access to compute, organisational adaptation, and effective data pipelines are becoming increasingly important for harnessing AI capabilities. At the same time, advantages such as population size may diminish in importance, while highly skilled talent becomes more valuable. The policy measures states adopt to navigate this shift are not chosen in a vacuum; they are justified, contested, and made legible through competing narratives.

The essay proceeds in three parts. First, we identify the five dominant narratives in AI geopolitics and the audiences each serves. Second, we map these narratives to specific instruments across the AI technology stack – compute, data, and models – to show how narrative choices generate concrete policy. Third, we propose a set of recommendations that serve India’s national interest.

Narratives Shaping AI Strategies and Governance

Narratives help societies make sense of technological change by translating developments into stories that people can relate to. They play a particularly important role in periods of uncertainty, when the capabilities, risks, and societal consequences of emerging technologies are yet to be fully ascertained.

Consider the argument that AI’s impact on the workforce will resemble that of past innovations like the internal combustion engine, which rendered some jobs obsolete but increased productivity and created many more jobs in the automobile manufacturing industry. This is one of many narratives on the impact that AI is likely to have on the workforce. It makes the case for embracing the technology regardless of the short-term disruption it is likely to cause, given its long-term effects of increased productivity and employment in the future.

With AI raising multiple competing concerns relating to its impact on society, climate, the economy, and the balance of power between nations, narratives play a critical role in influencing policymaking. They help frame priorities, shape public opinion, mobilise resources, and set the stage for policymaking. Policymakers frequently draw on narratives of prosperity, decline, hope, fear, and ambiguity to make sense of emerging technologies and their consequences. AI geopolitics is therefore characterised by competing narratives. We identify four prominent narratives below. 

AI as a Conventional Arms Race

The arms race narrative frames AI research and development as a zero-sum competition among states in which technological leadership confers decisive strategic advantages and falling behind rivals is unacceptable. This narrative serves the U.S. security establishment, frontier AI labs seeking national-security legitimation, and a broader China-hawk coalition. This framing has gained considerable traction in academic and policy discussions and popular discourse. This narrative rests on contested expectations that the exponential progress of AI systems seen in recent years will continue and eventually lead to superintelligence in the near future. Critics, however, argue that these assumptions reflect Silicon Valley hype cycles more than technical inevitability.

Several policy measures aimed at maintaining a lead over adversaries – which seemed unimaginable just a few years ago – have come into being. These include export controls on advanced semiconductors, high-bandwidth memory and advanced packaging tools. Further up the value chain, there is alignment by allies such as the Netherlands and Japan on export restrictions on lithography machines. Other legislative proposals include export control whistleblower programmes, restrictions on semiconductor investments in China, and certification requirements for advanced semiconductor exports. More broadly, this narrative leads to prioritising speed over other concerns, such as safety or ethics, in order to maintain technological leadership. 

AI as Nuclear WeaponsA more extreme form of the arms race narrative includes proposals to treat AI as similar to “catastrophic dual-use nuclear, chemical, and biological technologies”. This narrative also serves the interests of the U.S. security establishment and frontier AI labs seeking national-security legitimation.

Policy measures within this framing aim to pursue deterrence and limit proliferation at all costs, in addition to having a competitive advantage over rivals. Any state pursuing unilateral AI dominance is likely to be met with deterrence measures through kinetic or cyber means. Hendrycks et al introduce the concept of mutual assured AI malfunction (MAIM) resembling the logic of mutual assured destruction (MAD) that underpinned nuclear deterrence. Non-proliferation measures take the form of stringent controls on compute and models to ensure they do not fall into the wrong hands. Some of the more extreme proposals under consideration – such as hardware-based monitoring of AI chips and location tracking mandates – reflect a logic of containment and control closer to non-proliferation and distinct from that of the conventional arms-race narrative.

AI as a Geopolitical Innovation Race

In contrast to an arms race, an innovation race is a positive-sum game in which multiple actors compete and collaborate across national boundaries. Schmid et al. propose that a “geopolitical innovation race” more accurately describes the current situation, where great power politics and technology development are deeply entangled. They argue that this narrative is driven by a mix of motivations that include national security, economic leadership and national status. It involves a complex network of companies, states, and research institutions competing and collaborating on technology research and development, rather than a state-military nexus that would drive efforts in an arms race. As a result, it can generate both zero-sum and positive-sum outcomes for the actors involved. 

Policy measures associated with this narrative seek to maintain an edge in innovation while preserving global collaboration. They may include measures that limit a rival’s technological progress, such as export controls on inputs such as semiconductors. This framing can also justify subsidies and incentives for building data centres, as seen in the EU, the U.S., China and India, with the aim of making domestic companies more competitive on the global stage.

AI as a General-Purpose Technology

This narrative views AI not as a potentially superintelligent technology in the foreseeable future but as a transformative general-purpose technology akin to electricity or the internet. Under this framing, widespread adoption is likely to occur over decades and will be constrained by factors such as the capability-reliability gap, learning curves, organisational reforms, compliance with laws and norms, and plain inertia. This narrative places greater emphasis on safety and ethics than on the speed of development.

States should place greater emphasis on building the capacity to “diffuse”, or widely adopt, innovations rather than relying on protectionist measures to strengthen national innovation capabilities. Diffusion depends not just on the ability to innovate but also on institutions that promote technology transfer, human capital development, and openness to trade. 

AI as a Global Commons

This narrative is often promoted by international organisations and civil society groups that advocate for AI as a tool to benefit all humanity.  Several ethical arguments support this claim. One is that AI systems are trained on vast amounts of human labour and content, which could eventually impact livelihoods at scale. Another is that the large and growing energy demands of data centres are likely to have consequences on the environment, the costs of which will be borne disproportionately by the more vulnerable sections of society, while benefits accrue to a few.

While this narrative does not align with how incentives around AI development and deployment work, nation-states, particularly middle powers or low- and middle-income countries, use it as a diplomatic strategy. It offers an alternative to the current landscape in which the U.S. and China dominate technology research, development, and norm-setting. For example, it aligns with India’s diplomatic position in forums like the G20 or Global Partnership on AI (GPAI), where it has championed the idea of AI for all.

Narratives are not merely framing devices. They help build consensus, mobilise resources, and lay the groundwork for policymaking. Different stakeholders seek to shape the narratives around AI governance so they can shape the rules around its governance. A state that adopts the arms-race narrative reaches for compute denial and model export controls; one operating under the nuclear-weapons logic may add non-proliferation and hardware monitoring. The innovation-race narrative produces subsidies and ecosystem building, while the general-purpose technology narrative emphasises diffusion infrastructure, regulatory capacity, and human capital. The global commons narrative, in contrast, supports multilateral governance and redistribution mechanisms. The map below shows which narratives currently dominate each layer of the AI stack, and which layers are most contested.

The following section examines how geopolitics determines access across the key layers of the AI technology stack. 

Geopolitics Across the AI Technology Stack

The narrative a state adopts determines which layers of the AI stack it treats as strategic and which instruments it adopts. For instance, under the Biden administration, the United States adopted a “small yard, high fence” approach aimed at protecting sensitive technologies with high trade and investment controls. However, regardless of the administration, the “yard” has continued to expand in scale, even if enforcement has not always been effective. AI systems are built on key building blocks such as data, compute, models and applications. Across these building blocks lie bottlenecks that are being leveraged for geopolitical ends. 

Considering the entanglement of technology with national power, states are likely to pursue strategies aimed at securing their relative advantage in the AI domain over their rivals. Strategic objectives could take the form of denial, outpacing adversaries, and removing major bottlenecks. Table 1 outlines the key instruments of AI geopolitics and their often-overlooked repercussions. 

Table 1: Instruments of AI Geopolitics

Compute

Compute governance is influenced by the arms-race, nuclear-weapons, and geopolitical innovation race frames. Computing power is critical for the development and deployment of AI systems. Compared to other inputs, “compute is a particularly effective point of intervention: it is detectable, excludable, and quantifiable, and is produced via an extremely concentrated supply chain”, making it an attractive leverage point for governing AI.

For example, the Dutch company ASML has a monopoly on the production of EUV (extreme ultraviolet) machines required to fabricate advanced semiconductor chips. NVIDIA dominates the AI chips market and has become the most valuable company in the world, surpassing many established technology giants. This huge market share stems from its early investments in building GPUs and its widely used proprietary software interface, CUDA, which allows developers to utilise the hardware’s capabilities effectively. Further downstream, Microsoft Azure, Google, and Amazon Web Services together account for nearly two-thirds of the global cloud infrastructure market. These providers are the touchpoint for most users deploying AI at scale and leverage economies of scale and vertical integration to maintain their competitive advantage. These are just a few of the bottlenecks in the extremely concentrated supply chain.

Some U.S. policy measures deny adversaries access to compute capabilities. The now-rescinded AI diffusion framework, for instance, divided the world into three tiers based on levels of trust that determine access and security requirements for importing advanced AI chips and certain AI model weights. It has since been replaced by measures that strengthen chip-related export controls on China. Hardware-based mechanisms to monitor and restrict the usage of AI chips are also reportedly under consideration. In addition, the U.S. has introduced legislation mandating built-in location tracking for AI chips to prevent their diversion to China, Russia, and other countries of concern. Such measures are problematic, as they raise concerns surrounding ownership, privacy and surveillance. 

Strategic blocs also help build collective resilience. The Pax Silica is a U.S.-led strategic partnership initiative aimed at securing the global supply chain for AI across critical minerals, energy, advanced manufacturing, semiconductors, AI infrastructure and logistics. Key partners include the United States, Japan, South Korea, the Netherlands, the United Kingdom, Australia, Singapore, Israel, United Arab Emirates, Qatar, Sweden, and India.

Many countries are also building massive domestic data centres to incentivise domestic innovation and address sovereignty concerns. These include the EU’s AI factories, India's AI Compute Mission, and China’s state support for data centres.

Data

The narratives determining data governance are more mixed. Sovereignty and localisation are often associated with the innovation-race narrative, while open dataset initiatives gesture toward a commons-narrative. Data is an integral input to training AI. Large language models are trained on a vast corpus of human-generated content. As the supply of high-quality publicly available human-generated content for training AI is nearly exhausted, proprietary datasets with high information content and multimodal data sources such as images or video become increasingly invaluable. Technology platforms collect vast amounts of data in exchange for the free services they provide users, using it to feed their ad-based revenue models. In addition, data within walled gardens such as X and Reddit has become a valuable source of high-quality user data.

Data localisation requirements limit cross-border data flows due to privacy, security and sovereignty concerns. The EU’s General Data Protection Regulation, China’s Personal Information Protection Law, and India’s Digital Personal Data Protection Act, along with other sector-specific regulations, all require varying degrees of data localisation. This leads to increased compliance burdens due to varied regional regulations to house their data locally.

Governments are also creating datasets with positive externalities and security implications. Examples include the strategic and science datasets planned under the U.S. AI Action Plan and language datasets developed through initiatives like Bhashini.

Models

Narratives that determine model governance sit between innovation-race framing (efforts to promote domestic champions) and arms-race framing (restrictions on use cases and model weights). Most frontier AI models are designed by companies based in the U.S. or China. Models exist along a spectrum of openness. At one end are maximally open ones, where all model artefacts, including training data and code, are released under open licences. While these maximally open models are used primarily for open science, the less open models, with only model weights, documentation, and tools released under open licenses, also enable downstream innovation. They are particularly useful in applications where users may be unwilling to share sensitive data with a model vendor. Both China and the U.S. have clearly articulated in their national AI plans the importance of promoting the global adoption of their domestically developed open models, highlighting their strategic importance.

Model governance also involves restrictions on use. The now-rescinded U.S. diffusion framework also imposed controls on the model weights for certain advanced closed-weight dual-use AI models. At the same time, leading AI labs in the U.S. like OpenAI, Anthropic, and Meta have entered into deals that allow the government to use their models for military purposes. Since the terms of use for these models prohibit their use for security purposes by other nations, such arrangements raise questions about their implications on the balance of power between states.

Models also raise concerns of espionage and sabotage. These may include industrial espionage to gain access to proprietary data or model weights, which are not available in the public domain. Chinese AI labs such as DeepSeek, Moonshot, and MiniMax have been accused of distilling frontier models from American AI companies. In essence, this involves taking a stronger AI model’s outputs to teach a weaker model. The attacks were sophisticated and used deceptive techniques to mask the identity and intent of the distillers. The U.S. AI lab Anthropic claims that this happened on an industrial scale – “16 million exchanges with Claude through approximately 24,000 fraudulent accounts, in violation of our terms of service and regional access restrictions.”

Policy Implications

While most frontier research and development in the compute and model layers of the AI stack is being led by the U.S. or China, middle powers such as India have several pathways to navigate this landscape and strengthen their technology resilience and autonomy. These range from short-term measures to ensure access to the technology to longer-term measures aimed at bolstering the innovation ecosystem.

First, building a foundational level of base capabilities offers a pathway to developing knowledge and skills in critical areas over time. For example, India’s efforts to build semiconductor fabrication capabilities in 28nm chips aim to create an ecosystem around its existing strengths in semiconductor design capabilities. The compute, data, foundation model, and skilling pillars of the India AI mission seek to achieve a base level of capabilities and establish a domestic presence in various critical areas. However, with only 38 per cent of the 33,000 GPUs empanelled under the compute mission as of March 2026, adoption has been slow. Effective implementation will be key to ensuring that these initiatives achieve their desired outcomes. 

Second, funding and adopting open-tech ecosystems that can compete with proprietary alternatives in areas with high market concentration can help address vulnerabilities. These open-tech ecosystems need not be domestic and can still provide resilience due to their public-good nature. For example, developing open-source alternatives to CUDA, NVIDIA's proprietary programming platform, could help break the GPU stranglehold in AI. Ongoing initiatives such as OpenAI Triton, an open-source alternative to CUDA that allows AI models to run on diverse hardware, or Apache TVM, which optimises machine learning workloads across different processors, are promising alternatives. There are also lessons to be drawn from China’s open-source playbook, including active participation in open standard instruction-set architectures such as RISC-V, which allows companies to design chips without licensing fees or export restrictions.

While limited open-weight models, such as those released by NVIDIA, OpenAI, Meta, and DeepSeek, allow downstream uses, they do not provide the complete access required for further research and development. Supporting maximally open models, such as those developed by the non-profit EleutherAI, could generate positive spillovers across the research ecosystem. Alternatively, targeted government procurement or pilot adoption could also help create an early user base and support ecosystem development.

Such funding projects could also be sustained in partnership with states with similar goals, such as the EU, which recognises open-tech as a priority. Such collaboration could combine India’s software expertise and scale with the EU’s world-class research infrastructure. Creating an Open Source Programme Office that can crowdsource identifying projects globally and supporting them could further strengthen these efforts. A version of this model has been prototyped at a small scale by Zerodha’s Free and Open Source Software (FOSS) fund.

Third, an open-weight model pathway with domestic inference capacity. Models with near

frontier capabilities are currently released as open-weight models by leading American and

Chinese labs. India needs multiple domestic inference providers like OpenRouter that serve

multiple models rather than a single national AI cloud. Regulatory attention should shift from the

provenance of model weights to the deployment layer, which determines where data flows, who

controls the compute, and what audit trails exist. Given that the openness window may be

limited, the capacity to absorb, evaluate and fine-tune whatever is available has to be

developed. The GPUs running inference in India remain designed by American firms subject to

American export controls, which is a reason to continue engaging Washington through

arrangements like Pax Silica. A model’s refusals and framings are fixed in its weights –

domestic inference removes revocability risk, not values risk, which is why evaluation and red-

teaming capability must be developed.

Fourth, middle powers could explore asymmetric options to develop leverage over critical bottlenecks in the AI value chain. These could be niche areas in which a middle power can become indispensable and wield disproportionate power. Examples include the expertise developed by ASML and TSMC in critical parts of the semiconductor value chain. For India, one such opportunity may lie in building a strong talent base on open-source machine learning platforms such as PyTorch.

Fifth, while the measures outlined above can help build some short-term resilience, a broader innovation policy is required to strengthen the research and development ecosystem in academia and industry. A resilient future lies “not in a narrow focus on LLMs or GPUs, which are transient artefacts, but the societal and academic foundation required to enable conditions and ecosystems that lead to the creation of breakthroughs like LLMs.” Achieving this objective requires strengthening both innovation and diffusion capabilities. The Anusandhan National Research Foundation already has a mandate aligned with this goal, but its scope can be scaled further. It should be empowered to fund future-looking, high-risk, high-reward research that moves beyond traditional academic metrics. 

Finally, India can leverage its human capital as a diplomatic asset. The country is poised to become the world’s AI talent hub and can leverage this advantage to build talent partnerships with other countries and regions, such as the EU and Japan. 

The recommendations above – base capability building, open-tech investment, asymmetric leverage in niche bottlenecks, innovation ecosystem reform, and human capital diplomacy – should be read as the operational arm of the general-purpose technology or innovation race story at the domestic level, while the commons rhetoric continues to do work to build collaboration and broaden access at the global level. We argue that these approaches are not contradictory; rather, they reflect the deliberate deployment of different narratives at different levels to serve varying audiences and strategic interests. The scale of the initiatives under the IndiaAI mission remains modest in comparison to private sector investments in the U.S. or China. In the near term, India’s comparative advantage lies in enabling faster diffusion of the technology. The recommendations offered in this essay aim to support this objective while simultaneously laying the foundations for long-term capabilities and resilience.