-1793607240500.png)
The Ministry of Electronics and Information Technology (MeitY) has launched a handbook to help institutions share data securely for artificial intelligence (AI) training while protecting privacy and data ownership.
The Handbook on Data Empowerment and Protection Architecture (DEPA) for AI Training was unveiled by MeitY Secretary S Krishnan at a roundtable in Bengaluru on October 31. The initiative was developed with support from the iSPIRT Foundation.
The framework aims to address concerns around privacy, confidentiality, intellectual property and institutional control that often prevent organisations from sharing data for AI development.
How will the DEPA framework work?
The DEPA framework provides a legal and technical structure for institutions to collaborate on AI training without losing control over their data.
It combines contractual permissions, controlled computing environments and privacy-enhancing safeguards. These measures are intended to ensure that data is used only for approved purposes while protecting the interests of the organisations that provide it.
The framework also includes data provenance mechanisms. These allow the origin of data, permissions granted for its use and its role in the AI training process to be tracked.
This is expected to improve transparency and accountability in the development of AI models.
AI models to access data without transferring it
Principal Scientific Adviser to the Government of India Ajay Kumar Sood explained that the framework allows AI models to access data without requiring institutions to transfer their datasets.
“Under DEPA, the model travels to the data, rather than the data travelling to the model,” Sood said.
The approach is expected to support collaboration between multiple institutions, including those working across national borders.
It could also help researchers access diverse datasets for scientific studies and AI development while maintaining data protection measures.
Framework aims to support responsible AI development
MeitY said access to diverse and high-quality data is important for AI development. However, concerns about data security and ownership often limit collaboration between organisations.
The DEPA Training Framework aims to address these challenges by enabling trusted data sharing while protecting privacy, commercial confidentiality and institutional control.
S Krishnan said the framework would help unlock access to large datasets for AI training while ensuring that data is used responsibly and securely.
The Bengaluru roundtable brought together representatives from government, industry, technology, academia and research to discuss the role of secure data collaboration in supporting AI innovation.
The initiative is intended to strengthen India’s AI ecosystem by making data sharing more secure and transparent.
ALSO READ| BITS Pilani Scholarship 2027: 100% Tuition Fee Waiver for Top 1,000 BITSAT Rankers
