Talent-Rich, Compute-Poor: A Layer Decomposition of India's Position in the Global AI Race

Artificial intelligence is often talked about as a single, high-stakes race where nations sprint toward a finish line, with the United States and China currently leading the pack. To understand who is winning, experts rely on composite indices, which are like report cards that combine many different factors—such as how many researchers a country has, how much money is invested in new companies, and how powerful its computers are—into one single score. These scores help governments and businesses decide where to focus their efforts, but they can sometimes hide the messy reality underneath. A country might look like a top contender on a list because it has a massive number of skilled workers, even if it lacks the heavy machinery needed to build the most advanced systems. This creates a confusing picture for a country like India, which is frequently cited as a global powerhouse for engineering talent but rarely seen as a developer of the cutting-edge AI models that dominate headlines.

A new report by Alfredo Merlet seeks to clear up this confusion by looking past the final rankings to see exactly where India stands in relation to the world leaders. The author used two major, independent assessments of the global AI landscape: the Stanford Global AI Vibrancy Tool, which measures the overall energy and activity of a country's AI ecosystem, and the Oxford Insights Government AI Readiness Index, which evaluates how well a government is prepared to use these technologies. By breaking down the scores from these tools into their individual parts, the study reveals a stark and specific reality. India is not losing the race in every area; in fact, it is performing exceptionally well in the human element. However, the country is falling dramatically behind in the physical and financial resources required to train the most powerful AI systems.

The investigation began by taking the official scores that place India third in the world according to the Stanford tool and twenty-seventh according to the Oxford index, and then asking what those numbers actually meant. The researcher treated the gap between India and the United States like a puzzle, separating the total difference into distinct layers: the depth of the economy and the number of companies, the availability of computing power and infrastructure, the volume of research and new models, the quality of the workforce, and the quality of government policy. When these layers were weighed according to the original authors' own formulas, a clear pattern emerged. The massive difference between India and the United States is almost entirely driven by a lack of capital and computing power.

In the Stanford assessment, the gap in economic depth and computing infrastructure accounts for more than sixty percent of the distance between India and the United States. In the Oxford assessment, these same factors make up nearly fifty-seven percent of the gap. By contrast, the talent gap is almost non-existent. When looking at the availability of skilled workers, India's score is so close to that of the United States that the difference contributes less than one percent to the overall gap in the Stanford ranking. In the Oxford ranking, talent accounts for only about five percent of the distance. This means that India is not behind because it lacks smart people or engineers; it is behind because it lacks the expensive, energy-hungry hardware and the massive financial investment needed to run them.

The raw numbers behind these scores tell an even more dramatic story. In 2024, the United States possessed a computing capacity that was roughly 220 times larger than India's. While the United States had private investment in artificial intelligence totaling nearly 109 billion dollars, India's private investment was just 1.16 billion dollars. Furthermore, while the United States had developed 47 notable new AI models and China had developed 26, India had developed none that made the global list. Yet, on the human side, India's relative skill penetration and the growth rate of AI hiring were actually higher than those in the United States. The country has successfully built a world-class workforce capable of deploying and applying technology, but it has not yet built the industrial base required to create the frontier technology itself.

This imbalance has been shifting over time. Between 2017 and 2024, as India's overall ranking improved, the nature of the gap changed. The country has been closing the distance in knowledge and research output faster than it is closing the distance in assets. Consequently, the remaining gap is becoming more concentrated in the areas of capital and computing. The study notes that this is a stable configuration, consistent with how latecomers often enter new technological fields: they can leapfrog in layers with low entry barriers, such as education and services, but they struggle to break into layers that require massive scale and heavy investment, which are currently controlled by the established leaders.

The report also highlights that the famous "third place" ranking for India is somewhat of an illusion created by the way the data is measured. When the scores are adjusted to look at the population size rather than the total national output, India falls to the very bottom of the list of 36 ranked countries. This suggests that while India is a giant in absolute terms, its advantage is diluted when viewed on a per-person basis. The study argues that the current public programs in India, which aim to build a certain amount of computing power through public-private partnerships, are too small to close a gap that is 220 times wide. The entire five-year budget for the new AI mission is roughly equal to just one year of India's own private investment, and it represents only a tiny fraction of what the United States invests in a single year.

Ultimately, the research concludes that India is not losing the AI race across the board; it is simply running a different race. The country is strong in talent and in the ability to spread and use technology, but it is structurally constrained by a lack of frontier assets. The findings suggest that for India to truly compete at the very top, it would need to address the massive deficit in computing power and capital, a challenge that current policy measures are not yet large enough to solve. The country's strength lies in its people and its digital infrastructure, but without a corresponding leap in hardware and investment, it will remain a talent-rich, compute-poor nation, capable of using the best tools but not yet capable of building them.