Relying on foreign AI systems risks economic, cultural, and technological dependency; India must prioritize domestic innovation and research, say experts
India today stands at a critical juncture in the area of artificial intelligence. While the country is among the fastest adopters of AI in the world, it remains heavily reliant on technologies developed elsewhere. This paradox, experts warn, cannot persist if India seeks technological sovereignty.
Professor Balaraman Ravindran, Head, Wadhwani School of Data Science and Artificial Intelligence and Centre for Responsible AI at IIT Madras, emphasises: “Most of the AI systems powering today’s applications are still developed outside India. While our developers excel at building on top of these systems, the foundational layers remain largely foreign-controlled. If India is serious about technological sovereignty, that must change.”
Some early progress is visible. Startups like Sarvam AI are developing smaller, India-focused models capable of understanding local languages and contexts. “AI that cannot understand India cannot serve India,” Ravindran stresses. Yet these are only initial steps. Building frontier AI models demands long-term, high-risk investment—a domain where India has historically lagged.
Globally, breakthroughs have been driven largely by private capital willing to take long bets. Open AI, for instance, continues to invest heavily despite limited profitability. India’s ecosystem, however, is often oriented toward short-term returns, nine-month project cycles, and rapid deployment—conditions ill-suited for foundational technological innovation.
India has built a workforce adept at application development, system integration, and scaling digital products. However, fewer professionals are trained in core AI research, model architecture, or systems-level innovation. “We are builders—but mostly of what others have already built,” Ravindran notes.
This imbalance carries strategic risks. If access to global AI systems is restricted due to geopolitics or economics, India could scramble to catch up. Cultural consequences are equally significant. Many global AI systems misrepresent India, propagate gender biases, misinterpret geography, and overlook local health realities. “An AI system that does not understand India will not just fail users—it will misinform them,” Ravindran warns.
Even building domestic models is insufficient without hardware sovereignty. India remains dependent on imported GPUs, semiconductors, and cloud infrastructure. “Without sovereign silicon, sovereign AI remains incomplete,” Ravindran adds.
Unlike the US, China, or the EU, India’s emerging AI strategy balances innovation, regulation, and inclusion. Badrinarayanan Gopalakrishnan, economist and Distinguished Fellow, Pahle India Foundation, and former head, Trade, at NITI Aayog, explains: “India’s strength lies not in copying global AI models but in building systems that solve real-world problems for billions of people. A human-centric approach is not just ethical; it’s strategic.”
India has already demonstrated the power of digital public infrastructure with the Unified Payments Interface (UPI). The next frontier is AI as Digital Public Infrastructure, enabling startups, governments, and researchers to build scalable solutions for sectors like healthcare and agriculture. Initiatives like Bhashini—enabling voice interfaces in multiple Indian languages—showcase how AI can improve accessibility for low-literacy populations.
Yet structural bottlenecks remain. High-end compute is concentrated in global tech giants, local language models are scarce, and AI infrastructure is power-intensive. “Leapfrogging legacy systems allows India to innovate without being shackled by early-mover inefficiencies,” Gopalakrishnan notes.
Focusing solely on applications or the ‘diffusion layer’ of AI is a risky strategy. Rajat Sethi, Visiting Fellow at India Foundation and Co-founder of QuBeats, warns: “Focusing only on applications over foundational AI is a strategic gamble. Control over the full stack—from chips to models—is essential for sovereignty.”
India’s AI stack extends from critical minerals powering data centres to homegrown GPUs, energy sources, large models, and applications. Without investment in each layer, the country risks becoming a “digital colony”, where foreign AI giants exploit Indian data, talent, and markets. Sethi urges government action to create compute parks, enable energy availability, and incentivize domestic AI data centres.
He further emphasizes the need for private sector engagement: “Industrial houses must rethink R&D. Without meaningful investment, India cannot retain talent, stimulate MSME innovation, or compete globally. Public accountability—naming and shaming laggards—can catalyse change.”
Despite challenges, India’s AI journey presents unparalleled opportunities. Its young, digitally native population, coupled with agile policy, can nurture startups and innovation. “We must act now to build infrastructure and talent, or risk the AI revolution happening around India, not for India,” Sethi cautions.
Experts agree that India does not need to mimic Silicon Valley or Shenzhen. Its potential lies in building scalable, resource-efficient AI systems optimised for constrained environments—a unique advantage as sustainability challenges loom globally.
The choices India makes today—on investment, governance, and inclusion—will shape not just its economic trajectory but its strategic autonomy for decades. As Gopalakrishnan observes: “AI must serve society, not just markets. This is India’s opportunity to lead a paradigm shift from profit-first to impact-first innovation.”
The question is: Will India remain a passive participant in the AI revolution, or will it harness this epoch-defining technology to build a sovereign, inclusive future? The answer depends on the decisions made today.
