AI Power Is Splitting: America Dominates, China Cuts Costs, India Gains Ground

The United States remains the dominant force in artificial intelligence, supported by frontier AI labs, semiconductors, cloud infrastructure and investment strength. But China’s rapid gains in low-cost AI models and India’s rise to third place in a global AI readiness index suggest the next phase of competition will be shaped not only by technological leadership, but by affordability, deployment capacity and market depth.

The global artificial intelligence (AI) race is no longer only about who can build the most powerful model, but also about who can make AI cheaper, more widely deployable and more useful to large markets.

The United States still dominates the global AI ecosystem, backed by frontier laboratories, semiconductor strength, hyperscale cloud infrastructure and deep investment. However, China is becoming a formidable challenger in cost-efficient AI models, while India is emerging as one of the world's most AI-ready markets, trailing only the US and China in Stanford University's Global AI Vibrancy Index cited by J.P. Morgan.

The US remains the high-end leader. China is challenging the economics of AI adoption. India is gaining relevance as a major readiness and deployment market, even though it remains far behind both countries in overall AI capabilities and semiconductor ecosystem development.

America Still Controls the High Ground

The US remains the centre of gravity in artificial intelligence. J.P. Morgan describes it as "the most vibrant and prepared country for AI," with China close behind on some measures. The lead reflects the strength of America's technology ecosystem, where frontier AI companies, chipmakers, cloud providers and large enterprise users reinforce one another.

US companies continue to dominate the AI hardware stack. Nvidia accounts for the bulk of global AI accelerator revenues, while Google, Amazon, Microsoft and Meta are developing custom-designed chips to reduce costs and expand computing capacity. AI leadership depends not only on better software, but on access to the computing power needed to train, run and scale advanced models.

The US also has an advantage in productivity-linked AI adoption. J.P. Morgan notes that the US leads the G10 in labour productivity and total factor productivity, and says productivity growth in information and data-processing sectors accelerated significantly after the launch of generative AI tools. The point is important: AI's long-term value will be judged not just by benchmark scores, but by whether companies can turn it into measurable gains in output, efficiency and competitiveness.

For advanced uses, the American lead remains difficult to dislodge. Frontier AI models developed by US firms are still seen as essential for demanding applications such as cybersecurity, scientific discovery, agentic systems and large-scale reasoning. In those areas, capability, reliability and integration may matter more than price alone.

China Is Turning Cost Into a Weapon

China's challenge is not simply that its AI models are improving, but that some are becoming competitive at a fraction of the cost. J.P. Morgan says the "efficient frontier" in intelligence-per-dollar is "dominated by China," naming DeepSeek, MiniMax, Xiaomi and Alibaba. In comparative assessments of AI performance and operating costs, Chinese models increasingly occupy the segment where capability and affordability intersect.

By April 2026, leading Chinese open-weight models scored within a few dozen Elo points of closed frontier models and cost 10 to 50 times less per token, according to the report. Open-weight models give users more access to model parameters than closed systems, making them attractive to businesses that want lower costs, greater control and more flexibility.

The pricing gap is crucial because AI usage is moving from experimentation to routine deployment. When companies use AI occasionally, high per-token costs may be manageable. When AI is embedded into customer service, coding, internal search, summarisation, document processing and workflow automation, operating costs become harder to ignore.

J.P. Morgan also notes a surge in API calls to Chinese models on OpenRouter, suggesting growing market interest in lower-cost alternatives. It does not prove a wholesale shift away from US frontier models, but it does show that users are testing and adopting cheaper options where the cost-performance trade-off is attractive.

The result may be a split AI market. US frontier models could continue to dominate the most complex and sensitive tasks, while Chinese models capture more routine, high-volume and cost-conscious workloads. In that scenario, China does not need to overtake the US at the frontier to become a major force in enterprise AI. It only needs to make "good enough" AI much cheaper.

India's AI Moment Is About Readiness, Not Dominance

India is not yet a frontier AI power on the scale of the US or China, but its third-place ranking in Stanford University's Global AI Vibrancy Index gives it a stronger position in the global AI conversation.

The index measures countries across research and development, infrastructure, education, policy, governance and economic readiness. India's ranking suggests that it has many of the conditions needed to absorb and deploy AI at scale, even if it remains behind in advanced model development and semiconductor depth.

India's opportunity may lie less in building the world's most powerful AI models and more in becoming one of the world's most important AI adoption markets. Its technology services base, large digital economy and enterprise sector could make it a major user and implementer of AI systems across industries.

The capital-market angle strengthens India's relevance. J.P. Morgan notes that the 10 largest US stocks accounted for 17% of the S&P 500's market capitalisation in 2015 and now represent around 40%. Yet it adds that only Japan and India have lower equity concentration than the US by that measure.

AI boom has intensified attention on a small group of dominant US technology companies. India's lower equity concentration suggests a more diversified market structure. It does not mean India has a deeper AI ecosystem than the US, but it does make India relevant to investors looking at market breadth, diversification and long-term technology adoption.

The Next Battle Is Over Scale, Trust and Price

The emerging AI order is not a simple story of American decline or Chinese replacement - it's a story of different strengths. The US leads in frontier capability, chips, cloud infrastructure, investment and enterprise distribution. China is gaining ground in cost-efficient models that could win high-volume commercial workloads. India is building readiness and market depth, but still needs to close gaps in core AI capability and semiconductor development.

This creates a more complicated procurement decision for companies. The best model will not always be the right model. Enterprises will increasingly weigh performance against cost, data security, compliance, reliability, governance and vendor dependence. A cheaper model may be suitable for routine work, but not for sensitive or mission-critical uses. A frontier model may be powerful, but too costly for every task.

For investors, the question is where AI value will accumulate. If demand remains concentrated in advanced chips, cloud platforms and frontier systems, the US advantage may deepen. If lower-cost models expand adoption and reduce pricing power, value may shift toward companies that deliver efficiency at scale. India's market profile adds another layer by offering a less concentrated equity structure and a large potential AI adoption base.

AI is tied to productivity, industrial competitiveness, national security and global influence. J.P. Morgan warns that policy restrictions and supply-chain vulnerabilities could affect the future US lead, making semiconductors, export controls, cloud capacity and domestic AI investment central issues to watch.

The next phase of the AI race will be tracked through token pricing, benchmark performance, API usage, enterprise procurement, chip supply, cloud spending and national AI policies. The key question is whether China's cost advantage persists, whether US firms respond with cheaper and more efficient offerings, and whether India can turn readiness into deeper capability.