AI Kosh: India’s Unified AI Data Platform
The Indian government has extended an invitation to private companies, startups, and research institutions to contribute anonymous, non-personal datasets to AI Kosh. This initiative aims to provide high-quality data for training artificial intelligence (AI) models and accelerating AI advancements in India.
AI Kosh, launched under the IndiaAI Mission on March 6, integrates datasets from various sources, including government and private entities. With an allocation of Rs 199.55 crore, this initiative seeks to create a comprehensive AI dataset repository, enabling better AI research and development.
Why AI Kosh Needs Private Sector Participation
Data plays a crucial role in AI model training. While the government holds extensive datasets, contributions from private firms can add significant value. By inviting companies such as Google, Uber, and PhonePe to share anonymized usage patterns, AI Kosh aims to build robust AI models tailored to India’s needs.
A senior government official stated that many companies had shown interest in contributing to AI Kosh. To streamline the process, the government released a standard Expression of Interest (EoI) for all potential contributors. This ensures a structured approach instead of handling individual company requests.
Current Contributions and Collaborations
Several private firms and public institutions have already started contributing datasets to AI Kosh. Companies like Sarvam AI, Ola Krutrim, and Eka Care have shared their non-personal datasets. Government agencies have also provided valuable data, including census records, meteorological data, and industry-specific datasets from the ministries of agriculture and mines.
Additionally, AI Kosh has partnered with the Lok Sabha Secretariat and is in discussions with state broadcasters Doordarshan and All India Radio to share archives. Other contributors include Open Data Telangana, the Indian Council of Medical Research, and Digital India Bhashini Division. Research organizations like Development Data Lab and non-profits such as I-Hub for Robotics and Autonomous Systems Innovation Foundation have also joined the initiative.
What Type of Data Will Be Included?
The platform currently hosts 339 datasets and 159 AI models from 17 organizations across 15 sectors. Some notable contributions include:
- Text-to-speech and generative models in Indic languages from AI4Bharat (IIT Madras)
- Microsoft’s smaller Phi series of models
- Specialized non-LLM models
Datasets that can significantly impact AI development include healthcare data, doctor prescriptions, and call center interactions. Conversational datasets, in particular, will be valuable for training large language models (LLMs).
Ensuring Data Privacy and Compliance
AI Kosh ensures strict adherence to privacy laws. The shared data must comply with the Digital Personal Data Protection Act (DPDPA) and the National Data Sharing and Accessibility Policy (NDSAP). The government has clarified that AI Kosh will not engage in data monetization, reinforcing its commitment to ethical AI development.
All contributions must align with India’s data privacy regulations to ensure users’ anonymity and security. The government has assured companies that no personally identifiable information (PII) will be included in the datasets.
How AI Kosh Will Benefit India’s AI Ecosystem
The availability of diverse and high-quality data is essential for AI-driven innovations. By combining government and private datasets, AI Kosh aims to:
- Enhance AI research by providing India-specific datasets
- Improve the accuracy of AI applications across industries
- Boost AI model training for Indic language processing
- Foster collaborations between academia, businesses, and research institutions
With AI becoming a critical component of various industries, AI Kosh serves as a foundational step toward making India a global AI hub.
The Road Ahead for AI Kosh
The AI Kosh initiative aligns with India’s vision to lead in AI innovation. The platform will continue expanding its data repository by encouraging more organizations to contribute. Companies, startups, and research institutions can now take part in this initiative, shaping the future of AI in India.
As AI adoption grows, AI Kosh will play a crucial role in enabling businesses, researchers, and policymakers to leverage data-driven solutions. By ensuring accessibility, transparency, and compliance, this initiative is set to make a significant impact on India’s AI landscape.
Leave a Reply