India could be a different kind of AI superpower
It won’t look like America or China. It could still be a winner
Artificial intelligence (AI) is taking off in India. The country is now the second-largest market for OpenAI, whose ChatGPT service has 700m active users worldwide. Anthropic, another AI startup, also counts India as its second-largest market by usage. That reflects not just India’s huge population but also its appetite for new technology. According to BCG, a consultancy, 92% of Indian office workers regularly use AI tools, compared with 64% in America. In contrast to rich countries, a large majority of Indians believe AI’s benefits outweigh its risks.
The enthusiasm has been boosted by the “blitzscaling" tactics of Silicon Valley firms. OpenAI is selling access to a chatbot in India for a fifth of the price of its cheapest American plan. Grok, from Elon Musk’s startup, xAI, is priced at a quarter of its American rate. Perplexity, a generative-AI upstart, has made its service free for a year to all 360m users of Airtel, a big Indian mobile operator.
Even in India, however, the boom in AI causes anxiety. The country’s youth-unemployment rate stands at 16%. Jobs in manufacturing have grown far less than had been hoped, in part because of rising automation—a trend that AI threatens to accelerate. White-collar work is starting to look wobbly, too. Tata Consultancy Services, India’s biggest IT-services firm, recently said that it would cut 12,000 staff in order to become “future ready". Jefferies, an investment bank, predicts that more IT firms will follow.
A second fear is foreign domination. Some Indian policymakers and investors fret that their country will come to rely on AI products and services that are controlled by firms from abroad. Keen not to be left out of the model race, the government has tapped Sarvam AI, a local startup, to build India’s first home-grown foundational model.
Such worries are understandable. Yet India stands to gain far more by embracing AI than it will lose. Many benefits of being open to global tech firms are already plain. The country’s payments network, which handles around 700m transactions a day, is piloting the use of AI to spot fraud in real time. Deeper transformations may emerge. AI assistants might help mitigate India’s chronic shortages of teachers and doctors.
Foreign dependency is also less of a problem than it may seem. True, India does not create the latest models or the fastest AI chips. But its firms can innovate in a distinctive way, by turning AI into world-beating products and services. India has the world’s second-largest pool of developers on GitHub, a coding platform, and a vast domestic market in which global tech giants and local firms compete side by side. That gives its firms both the talent and the testbed to create usable, affordable services that do the sorts of things that ordinary people want from AI.
Already, Indian users are shaping how the most popular AI models develop. Voice, not text, has quickly emerged as the dominant way of interacting with AI tools in India, in part because some users struggle to read. Indian firms are especially adept at designing services for a varied audience.
The “India Stack"—India’s digital platform for biometric identification and payments—has already become a model for other countries. Products infused with AI could be the next export of this kind: frugal, scalable innovation that is pioneered in India but adopted across the developing world. India’s path will not mirror America’s or China’s. But it could prove no less consequential. For billions in poorer countries, the shape AI takes could depend on what happens in India.
Subscribers to The Economist can sign up to our Opinion newsletter, which brings together the best of our leaders, columns, guest essays and reader correspondence.
topics
