New Delhi: Days after the launch of the low-cost foundational AI model DeepSeek by a Chinese startup, India’s government has announced plans to develop its own large language model under the Rs 10,370 crore IndiaAI Mission, IT Minister Ashwini Vaishnaw said on Thursday.
This initiative aims to reduce India’s dependence on foreign AI models, ensuring the technology is free from biases and accurately represents the nation’s diversity.
To achieve this, the government is establishing a large-scale AI infrastructure, including a facility equipped with 18,000 GPUs—essential for AI training and development. DeepSeek’s success, despite US sanctions limiting its access to GPUs, has sparked discussions about whether companies like OpenAI and Google are over-investing in AI development when a more efficient, cost-effective approach might be viable. However, big tech firms continue to reject this notion and remain committed to heavy AI investments.
India’s planned facility, expected to be operational within the year, aims to develop an indigenous large language model within ten months. This model will be designed to understand and respond effectively to the nuances of Indian languages and cultural contexts.
The initiative is a crucial step in India’s strategy to establish itself as an AI powerhouse while reducing reliance on foreign technology. Additionally, the project includes AI data centers in Odisha, prioritising data security and fostering innovation across multiple sectors.
By developing its own AI, India can enhance privacy, improve support for local languages, and maintain greater control over AI applications within the country.
