India needs to build its own artificial intelligence (AI) models as there is a risk that open-source ones may not remain available in the future, even though some industry experts have suggested India should just focus on building AI applications, according to Ministry of Electronics and Information Technology (MeitY) Secretary S Krishnan.
In an interview with Outlook Business, he also said that the IndiaAI mission includes both the development of applications and the creation of foundational models.
Open Source technology refers to the technical tools whose source code is available for developers to use or modify. While LLMs are the computer programs that use machine learning to interpret human language; Claude and DeepSeek are famous examples of LLMs.
Recently, Vaishnaw also announced that the government's IndiaAI Mission has successfully empaneled around 19,000 graphics processing units (GPUs) from various data centres.
The IndiaAI mission has invited proposals from startups, researchers, and entrepreneurs to collaborate on developing advanced foundational AI models trained on Indian datasets. This initiative aims to create indigenous AI models that meet global standards while addressing India's unique challenges and opportunities.
LLMs, which form the backbone of generative AI applications, require vast computational power and large-scale datasets for training. By leveraging these GPUs, IndiaAI aims to create AI models that can drive advancements across sectors.
As India empaneled 19,000 GPUs as a crucial step in its AI journey, the IT Minister said that on the line to democratise the technology, India’s AI mission is constantly working.
With the coming of DeepSeek from China has also defied the perception building LLMs cost billions of money. Therefore, the Indian government has interlinked the aspiration of establishing the country’s own AI foundational model with the idea of democratising the technology.
Another reason for why India desires to bring its own foundational model is because most of the AI models trained on Western data often come with inherent biases that has been mentioned by IT secy as well.
Further, IT secy also highlighted that India's annual spend on research and development (R&D) has remained stagnant because there is the lack of risk-taking behaviour among Indian companies, developing products involves high risks—while it offers the potential for significant profits too.
