News Brief

India Unveils 'AI Kosha' With 316 Datasets To Assist Indigenous Artificial Intelligence Model Development

Swarajya StaffMar 07, 2025, 01:22 PM | Updated 01:22 PM IST
Union Minister of Communication, Electronics and Information Technology, Ashwini Vaishnaw

Union Minister of Communication, Electronics and Information Technology, Ashwini Vaishnaw


The Union government on Thursday (6 March) unveiled 'AI Kosha', a dedicated platform for non-personal datasets, marking a major step in India's efforts to accelerate Artificial Intelligence (AI) research and development, reported The Hindu.

The initiative, a key component of the Rs 10,370 crore IndiaAI Mission, aims to provide structured data to help train AI models, particularly in Indian languages.

At launch, AI Kosha hosts 316 datasets, with a significant portion focused on language translation tools for Indian languages.

Other available datasets include health data from Telangana’s open data initiative, 2011 Census data, satellite imagery from Indian satellites, meteorological and pollution data

This platform is part of the broader IndiaAI Datasets Platform, which is one of the seven pillars of the IndiaAI Mission, a state-backed initiative to advance AI capabilities in India.

Union IT Minister Ashwini Vaishnaw, while announcing AI Kosha, provided updates on India's AI infrastructure.

He revealed that 14,000 Graphics Processing Units (GPUs) have now been commissioned for pooled access, up from the 10,000 announced earlier this year.

The Minister also highlighted the growing interest from startups in developing India’s own foundational AI model, a project that has gained momentum following the success of China’s DeepSeek.

“Now, the team is actually inundated with how to evaluate these applications,” Vaishnaw said, signaling a surge in AI startups keen on leveraging the new datasets and computing resources, quoted as saying by The Hindu.

AI Kosha is the latest in the government’s efforts to aggregate and make public datasets accessible for research and innovation.

The Open Governance Data Platform (data.gov.in) already hosts over 12,000 datasets from various government agencies.

To further encourage data sharing, Chief Data Officers have been designated across ministries and departments, tasked with expanding data availability for researchers, startups, and enterprises.

Join our WhatsApp channel - no spam, only sharp analysis