India's AI Crossroads: Service Economy to Imagination Economy

India has been referred to as the back office of the world over many years. Our IT industry now employs more than 5.4 million people, and contributes close to 10% of GDP and is projected to rise further by FY27. That is no small achievement. However, now that artificial intelligence is altering the manner in which technology operates in every place, we have a greater question. Are we still a service economy or are we another thing, an imagination economy?

 One can simply look at Silicon Valley and believe that we have to imitate them. Large scale, rapid growth, venture capital in search of the next app. However, imitation is no leadership. According to a NITI Aayog-McKinsey report, ‘AI for Viksit Bharat’, India can increase its GDP by up to 8% using AI, but only when we apply it in agriculture, healthcare, and education. That is the point. We cannot afford to repeat western problems but solve Indian ones.

Take healthcare. India allocates only 2.4% of GDP on health. That is abysmally small compared to what other countries spend. AI can help fill the gap. A NITI Aayog paper, ‘Digital India Beyond 2025’, discusses vernacular-first platforms. Suppose an AI is conversing in Marathi or Bhojpuri, and it is instructing a villager on how to treat tuberculosis. That is useful. That is leadership. Not another shopping chatbot.

 Agriculture is another area. Crop disease forecasting and soil analytics is an urgent need, with 65% of Indians still residing in rural regions. According to KPMG’s 2025 Tech Pulse report, only 6% of Indian technology companies patent annually, as compared to 22% in China. When our startups are agritech IP, food security and exports can be altered. This is not about catching up. It is concerned with agenda setting.

 Education too. The Union Budget 2026-27 renewed allocations of Rs 500 crore for AI in education and Rs 2,000 crore for AI infrastructure. The actual prospect lies in personalised learning of 250 million schoolchildren. Localised AI tutors may democratise education. In this case India can be a pioneer, provided we plan for diversity and scale.

 We are also not like the others. The Fifth Generation Computer Systems project in Japan in the 1980s did not succeed due to excessive central planning. The internet in the US was successful since DARPA succeeded in America because the government constructed the rails and allowed the trains to be operated by the private innovators. Taiwan created semiconductors using patient capital and long-term assistance. With UPI and India Stack, India demonstrated that open infrastructure is effective. The lesson is clear. We should not be as rigid as Japan, catalytic like DARPA, and patient like Taiwan, while remaining true to ourselves.

 The change should also be in the private sector. There are over 100 unicorns in India. However, fewer than 10% of the funded startups solve rural issues. It will be based on daring investments in native solutions.

 The AI crossroads in India is about choosing a path. We are coding someone else’s destiny. Or we can take imagination back, and create AI that speaks everything us. The decision will determine whether India will be remembered as the back office of the world, or the imagination economy that re-established AI leadership.

 (The author is a C-suite+ and startup advisor, and researches and works at the intersection of human-AI collaboration. Views are personal.)