India AI regulatory and data framework development

The Indian government is advancing regulatory and technical structures for artificial intelligence. Initially, the government considered a safety framework for frontier AI models, which may include a two-track system to distinguish standard applications from high-capability models. Proposed measures for these powerful models include pre-release evaluations, controlled-environment testing, and supervised sandbox requirements. Union Electronics and IT Minister Ashwini Vaishnaw indicated that a consultation paper would be released to gather stakeholder feedback.

Following these considerations, the Union Ministry of Electronics and Information Technology (MeitY) launched the Data Empowerment and Protection Architecture (DEPA) handbook for AI training. This framework establishes a techno-legal structure for purpose-bound data sharing, utilizing contractual permissions and privacy-enhancing safeguards. A central feature of the DEPA architecture is that the AI model travels to the data rather than the data being transferred, which is intended to facilitate secure multi-party and cross-border collaboration while ensuring data provenance and accountability.