The Geopolitical Implications of the Looming AI Bubble

Global spending on artificial intelligence (AI) is projected to reach US$ 2.52 trillion in 2026, with investments heavily concentrated in capital-intensive infrastructure and compute capacity. Microsoft, Alphabet, Amazon, and Meta alone are expected to commit between US$ 635 and US$ 665 billion, while Oracle has disclosed a five-year US$ 300 billion compute deal beginning in 2027. On the supply side, Nvidia is providing US$ 100 billion to OpenAI, and Amazon has pledged US $8 billion to Anthropic. These figures reflect a surge of capital increasingly driven by competitive urgency and speculative expectations as much as by demonstrated returns, raising concerns that valuations across the AI ecosystem are outpacing underlying profitability.

The trajectory of the investments within the AI sector mirrors the classic dynamics of an economic bubble. Intense competition among firms to secure technological leadership has fostered rapid capital deployment, reinforced by investor enthusiasm and fear of missing out. Media amplification and optimistic projections of transformative productivity gains have further accelerated this momentum, encouraging firms to scale infrastructure and acquire compute at unprecedented levels, often financed through aggressive spending commitments. However, as with past speculative cycles, this expansion depends on the realisation of expected growth rates and revenue streams that remain uncertain.

Should the sector fail to deliver commensurate returns, the current phase of exuberance could give way to distress. Slower-than-anticipated adoption, diminishing marginal gains from scaling, or constraints in monetisation may prompt reassessment among investors and firms alike. In such a scenario, the unwinding of positions could be swift, leading to sharp corrections in valuations, reduced capital expenditure, and broader financial strain across interconnected industries.

Assessing the Drivers and Durability of an AI Bubble

The paramount questions in this domain concern whether the contemporary AI market constitutes an economic bubble and, consequently, the likelihood of its rupture. As of October 2025, US equity analysts at Goldman Sachs contended that AI had not yet formed a bubble, pointing to public market valuations and transaction volumes that remained subdued relative to prior episodes. This assessment was reinforced by the robust performance of the ‘Magnificent Seven’ US technology firms (Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia, and Tesla), which continued to generate substantial free cash flows, execute share buybacks, and distribute dividends. Nonetheless, they cautioned that the escalating dependence of these corporations on debt to finance AI ventures posed a material risk to macroeconomic stability. Sam Altman, Chief Executive of OpenAI, the parent company of ChatGPT, has similarly observed that investors are exhibiting excessive enthusiasm for the current phase of AI advancement.

Despite these perspectives, the pronounced overvaluation of technology equities, coupled with the unpredictability of market volatility it induces, renders the emergence of an AI-centric economic bubble highly plausible. Although market capitalisations of AI stocks appear to rise steadily, this sustained increase is predicated on perceptions of enduring viability. This trajectory likely reflects a self-reinforcing feedback loop, in which rising prices attract further investment, which in turn drives prices higher.

The formation of an economic bubble does not inherently imply permanent, irrevocable damage to an economy. The optimal strategy for managing such situations is to sustain the bubble until equilibrium is restored. While this scenario is difficult to navigate and requires substantial additional capital, it prevents the ensuing sharp economic decline and low-risk appetite conditions that typically follow a bubble’s burst. China exemplifies this approach, exhibiting most symptoms of an economic bubble in the 2000s yet averting financial catastrophe by prolonging expansion until stabilisation occurred.

Major AI breakthroughs warrant vigilant monitoring, with controlled dissemination to safeguard returns on infrastructure investments.

Sustaining such a bubble necessitates circumventing critical pitfalls. Foremost, mitigating interdependence among industry leaders is essential, as it could trigger cascading failures. This requires diversification beyond the current oligopoly of dominant firms. Secondly, inadequate regulatory oversight and inconsistent government policies must be avoided. The US is performing well on this front, as the government has demonstrated proactive support for research and development. Finally, a counterintuitive measure involves deliberately tempering disruptive innovations that might prematurely render current infrastructure obsolete before sunk hardware costs are recouped. Major AI breakthroughs warrant vigilant monitoring, with controlled dissemination to safeguard returns on infrastructure investments.

US–China Strategic Competition in the AI Race

The US–China geopolitical rivalry over AI hardware and strategy significantly fuels concerns about an emerging AI economic bubble, exposing acute vulnerabilities in supply chains, investment risks, and uncertain paths to AGI. The global AI landscape faces a critical impasse, with China dominating the mining of rare earth minerals essential for manufacturing graphics processing units (GPUs) and other hardware vital for training AI models, while the US hosts leading designers and producers, notably Nvidia. This bifurcation has intensified the geopolitical race between the superpowers.

The US and China diverge markedly in their strategic prioritisation of AI. The American AI Action Plan underscores a resolute commitment to substantial investment, prioritising dominance in this domain even if it entails lags in other technological arenas. The United States (US) is so intent on this path that it aims to be the first nation to develop artificial general intelligence (AGI), which would confer an overwhelming strategic advantage on the global stage. Conversely, China pursues a more balanced approach, allocating resources to AI alongside other emerging technologies. Should an AI-centric economic bubble inflate and subsequently burst, the US risks ceding significant ground to China.

The US–China geopolitical rivalry over AI hardware and strategy significantly fuels concerns about an emerging AI economic bubble, exposing acute vulnerabilities in supply chains, investment risks, and uncertain paths to AGI. 

The central pillar of the current US strategy is heavy investment directed toward vast data centres that host immense computational capacity. This addresses the principal perceived barrier to AGI: insufficient compute scale. Between 2024 and 2025, Google, Amazon, Microsoft, and Meta spent approximately US$ 750 billion on such infrastructure to power their AI models, with projections estimating an aggregate of US$3 trillion invested in data and compute by 2029. These investments, however, carry manifold risks. They demand exorbitant upfront capital for asset-intensive builds, relying on specialised chips from firms like Nvidia, which rapidly become obsolete due to continual innovation. Entire data centres could become redundant should breakthroughs emerge in optimising models or alternative compute paradigms.

Compounding this is the mounting scepticism surrounding the viability of the prevailing US AI paradigm, centred on scaling and innovating Large Language Models (LLMs) to achieve AGI. Researchers have demonstrated that current models falter in genuine logical reasoning and rational cognition, merely simulating these through pattern-based predictions derived from training data. Consequently, this trajectory appears unlikely to deliver AGI, further casting doubt on the rationale for current investments and the inflated valuations of technology equities.

Collectively, these vulnerabilities suggest that US ambitions rest on precarious foundations. An AI bubble collapse could destabilise the economy, enabling China to supplant it as the preeminent technological power. Sustaining such a bubble until equilibrium, however, could secure enduring US ascendancy over decades.

Consequences of an AI Bubble Burst

Should an AI-centric economic bubble inflate and subsequently burst, the United States would bear the brunt of the fallout, furnishing China with a critical window to entrench itself as the preeminent technological superpower. This realignment would have profound implications across financial, technological, and geopolitical-strategic domains, underpinned by the axiom that technological supremacy buttresses military prowess, economic influence, and soft power, enabling China to shape international norms and institutions in its image.

• Financial Implications

A bubble collapse could catalyse profound economic disruptions, contingent on the severity of US impairment and China’s opportunistic gains. A protracted recession might accelerate de-dollarisation, as nations diversify reserves amid perceptions of US fiscal vulnerability. International organisations such as BRICS‑led initiatives have already pushed forward alternative payment systems, leading to visible impacts such as the use of the RMB in the total value of money flowing in and out of China rising in the first eight months of 2024; cross‑border RMB transactions increased by 21.1 percent compared with the same period in 2023. Power structures in global commerce could realign, positioning China at the epicentre through its Belt and Road Initiative (BRI) and dominance in critical supply chains, while institutions such as the World Trade Organisation witness diminished US influence, fostering Sino-centric trade paradigms.

• Tech Sector Implications

The US tech sector, heavily reliant on AI infrastructure investments, would face cascading failures from asset write-downs and eroded investor confidence. Dominant firms such as Nvidia, Google, Amazon, Microsoft, and Meta, having poured hundreds of billions into data centres and specialised chips, risk obsolescence amid supply chain frictions and paradigm shifts, amplifying market volatility. This vulnerability could fracture innovation pipelines, ceding proprietary advancements in AGI to China and precipitating a broader contraction in Silicon Valley’s oligopolistic dominance.

• Geopolitical Implications

The military consequences of this shift would include heightened tensions over Taiwan and the South China Sea, as a diminished US deterrent emboldens Chinese territorial assertiveness, generating cascading security dilemmas across South Asia and the Indo-Pacific. Existing international groupings, such as the Quadrilateral Security Dialogue (Quad), encompassing the US, Japan, India, and Australia, could become vulnerable to fracture, as European and Indo-Pacific partners would need to pivot towards pragmatic engagement with China. The AUKUS pact, too, could falter amid eroded US credibility, further unravelling Western-led security architectures.

Conclusion

In summary, the formation of an economic bubble around AI technology stocks appears increasingly probable, if not already underway. To avert a profound geopolitical realignment, the US must sustain this bubble until equilibrium is restored. Failure to do so risks China supplanting the US as the preeminent technological superpower, precipitating shifts in global paradigms, the erosion of existing alliances, and heightened security risks across the Indo-Pacific and South Asia.

This commentary originally appeared in Observer Research Foundation.