Delhi : India has emerged as one of the world’s most important markets for artificial intelligence. More than 100 million people in India use ChatGPT every week, and the country has a rapidly expanding AI-skilled workforce and developer community. The ₹10,372-crore IndiaAI Mission has strengthened this momentum. Yet becoming a major consumer of AI is not the same as becoming a leading producer of AI technology. Between India’s ambition and genuine leadership lie structural challenges involving computing power, talent, data, research, governance, energy and employment.
Access to computing power remains a major constraint. Under the IndiaAI Mission, more than 38,000 GPUs had been onboarded to the common compute facility by March 2026, and shared capacity had expanded beyond 45,000 GPUs by June. Yet India’s AI compute remains far smaller than that available in leading AI ecosystems.
The challenge is not simply GPUs. India must secure advanced processors, high-bandwidth memory, networking equipment and reliable electricity. US export-control policies also affect access to advanced AI chips, creating an additional strategic constraint. India is building a domestic semiconductor ecosystem, but it does not yet have large-scale domestic production of cutting-edge AI accelerators. India needs an entire ecosystem, not merely more imported GPUs.
India’s AI talent story is equally paradoxical. The country has a huge technology workforce and expanding developer community, yet it lacks sufficient numbers of frontier researchers, machine-learning specialists, semiconductor designers and AI infrastructure engineers. Industry estimates project demand for roughly 2.3 million AI professionals by 2027 against a talent pool of about 1.2 million, implying a potential gap of approximately 1.1 million. This is a projected mismatch, not a prediction of 1.1 million vacancies.
NITI Aayog’s 2025 roadmap points to uneven access to computer-science education in Indian schools. Unlike China and Russia, where computer science is taught more systematically, Indian students do not have uniform early exposure. NITI Aayog, drawing on Stanford’s 2025 AI Index, notes that India’s share of granted AI patents declined from around 8–10 percent of the global total in 2010 to below 5 percent in 2023. India’s research capacity must grow alongside its technology workforce.
Data presents another challenge. India has 22 constitutionally recognised languages and a vast range of linguistic varieties. Building high-quality, representative datasets across this landscape is difficult. Although frontier models have improved in several Indian languages, English still dominates much high-quality digital material. India’s task is not merely to translate English-language AI, but to develop datasets, benchmarks and models that understand Indian languages and contexts.
Initiatives such as Bhashini and the IndiaAI Mission are addressing this gap. The IndiaAI Mission has selected projects for indigenous foundation-model development, including large multimodal and smaller language models. The opportunity is to turn linguistic diversity into strength.
India has chosen a relatively flexible approach to AI regulation. In November 2025, the Ministry of Electronics and Information Technology unveiled the India AI Governance Guidelines, presenting a framework for responsible AI adoption. Built around seven principles and six governance pillars, the framework emphasises human-centred AI, innovation, safety and accountability.
The government does not impose a comprehensive standalone AI law, instead relying on existing laws and proposing mechanisms such as sandboxes. This flexibility may encourage innovation, but it raises a crucial question: who is accountable when AI causes serious harm? AI can affect employment, financial decisions, privacy, access to services and the information environment. Deepfakes can facilitate fraud and political manipulation, while algorithmic bias can reproduce historical inequalities.
Its implementation is phased, so it is inaccurate to suggest that the entire Digital Personal Data Protection framework is inactive until May 2027. The priority is to ensure that data protection and AI governance develop together, providing safeguards without unnecessarily restricting innovation.
Large AI-oriented data centres can require tens of megawatts of power and, at the largest scale, around 100 megawatts or more. India’s data-centre industry is expanding rapidly, but electricity generation, transmission, cooling and water infrastructure must expand alongside it. This matters because some technology centres already face water and power pressures. India’s AI strategy must include renewable energy, efficient cooling, water conservation and grid expansion.
Employment is perhaps the most consequential social challenge. AI exposure does not automatically mean job destruction. The IMF estimates that about 26 percent of Indian workers are in occupations highly exposed to AI. Some jobs are likely to benefit from AI and become more productive, while others face greater displacement risks. The questions are how many jobs will change, how quickly new jobs will appear and whether workers can acquire skills for expanding occupations.
NITI Aayog’s 2025 roadmap cites research suggesting that more than 60 percent of formal-sector jobs could be susceptible to automation by 2030, with IT and BPO among the exposed sectors. Around 400 million Indians work in the informal sector and have limited access to formal training. India therefore needs large-scale reskilling and AI literacy, not only elite training for researchers.
India has made investments in AI infrastructure, but frontier leadership requires greater research capacity and long-term capital. The first three AI Centres of Excellence received ₹990 crore over five years, while IndiaAI supports indigenous foundation models. Indian technology start-ups raised about $9.1 billion in 2025, including approximately $2.3 billion for deep-tech companies. These are significant achievements, but Indian AI companies still have access to less capital than the largest AI ecosystems.
India’s advantages are real: a huge technology market, a young workforce and a growing developer community, strong IT capabilities, digital public infrastructure and an ambitious AI programme. Aadhaar, UPI and Bhashini demonstrate India’s ability to build digital systems at enormous scale.
But adoption alone will not make India an AI superpower. India must expand advanced computing, strengthen semiconductor capabilities, develop research talent, create better datasets, finance long-term innovation, prepare workers for changing occupations and build governance systems that protect citizens without unnecessarily slowing progress. The decisive question is whether it can build the complete ecosystem required for sustained leadership at the technological frontier.
