India Aims for AI Breakthrough – But Is It Keeping Up?

21.02.2025

Two years after ChatGPT’s global breakthrough, China’s DeepSeek has disrupted the tech industry by significantly reducing the cost of developing generative AI applications. However, as competition for AI dominance intensifies, India seems to be lagging, particularly in building its own foundational language model for powering AI tools like chatbots. The government insists that a homegrown alternative to DeepSeek is on the horizon, providing startups, universities, and researchers with high-end chips to accelerate development within ten months.

Despite initial skepticism, global AI leaders are increasingly recognizing India’s potential. OpenAI CEO Sam Altman, who was previously dismissive, now states that India should play a leading role in AI. OpenAI’s second-largest user base is in India, while Microsoft has committed $3 billion to AI and cloud infrastructure. Nvidia’s Jensen Huang has also highlighted India’s technical talent as a crucial asset for the country’s AI future. With 200 startups focused on generative AI, entrepreneurial momentum is growing, but experts warn that India risks falling behind without structural reforms in education, research, and policy.

China and the US have a significant head start, having invested heavily in AI research, academia, and applications in military and law enforcement. India ranks in the top five on Stanford’s AI Vibrancy Index, which evaluates patents, funding, policy, and research, yet it remains far behind the two superpowers in key areas. Between 2010 and 2022, China and the US accounted for 60% and 20% of global AI patents, respectively, while India secured less than 0.5%. AI startups in India also received only a fraction of the private investment that their US and Chinese counterparts obtained in 2023.

India’s government-backed AI initiatives pale in comparison to global efforts. The country’s $1 billion AI mission is dwarfed by the US’s $500 billion Stargate project and China’s $137 billion AI strategy. While DeepSeek has shown that AI models can be developed using older, cost-effective chips, India’s lack of long-term capital from both the government and private sector remains a critical obstacle. Additionally, the absence of high-quality, India-specific datasets in regional languages like Hindi, Marathi, and Tamil poses challenges for training AI models suited to the country’s linguistic diversity.

Despite these setbacks, India stands out in AI talent, contributing 15% of the global AI workforce. However, many of these professionals are leaving the country due to limited research opportunities and a lack of deep-tech innovation in academia and corporate sectors. Experts argue that India’s AI development could benefit from the same collaborative model that drove its digital payments revolution. The success of the Unified Payments Interface (UPI), which transformed financial transactions through government-industry-academia cooperation, serves as a blueprint for fostering AI growth in India.