India Launches Data Framework to Secure AI Training Collaboration

The Ministry of Electronics and Information Technology (MeitY) has introduced a handbook detailing the Data Empowerment and Protection Architecture (DEPA) to facilitate secure data sharing for artificial intelligence development. IT Secretary S. Krishnan released the document during a roundtable discussion in Bengaluru on Saturday.

The framework establishes a techno-legal structure designed to allow institutions to collaborate on AI training datasets while maintaining control over their proprietary information. According to the ministry, the system utilizes contractual permissions, privacy-enhancing safeguards, and controlled computing environments to protect data integrity.

Ensuring Accountability and Traceability

The DEPA framework includes mechanisms for data provenance, which allows organizations to track the origin, permissions, and usage history of datasets throughout the training lifecycle. Officials stated this feature is intended to support transparency and accountability in AI development processes.

Principal Scientific Adviser Ajay Kumar Sood noted that the architecture operates on a principle where the AI model is brought to the data, rather than transferring the data itself to the model. This method is intended to enable multi-party and cross-border collaboration while addressing concerns regarding intellectual property, confidentiality, and institutional control.

Global Scientific Contribution

The ministry developed the handbook with support from the iSPIRT Foundation. Government representatives, industry leaders, and academic researchers attended the unveiling to discuss the implementation of these standards. Krishnan described the initiative as a step toward matching India’s AI ambitions with secure and responsible data practices. Sood characterized the framework as India’s contribution to international science diplomacy.