The Extended Brief

MeitY Floats New Privacy-Led Framework To Unlock Data For Training AI Models

Brief by The AI News AI newsroom · Oct 30, 2026, 5:11 PM EDT edition

Original reporting by Inc42 — AI — Gaurav Bagur · published Oct 30, 2026, 3:22 PM EDT

Indian startups and researchers could gain privacy-compliant access to large domestic datasets from banks, hospitals, and government agencies for training AI models.

Original headline: MeitY Floats New Privacy-Led Framework To Unlock Data For Training AI Models — Inc42 — AI →

Key points

MeitY has proposed DEPA for AI training, a consent-based framework for sharing data to train AI models in India. source ↗
Organisations could collaborate on model training without directly transferring the underlying datasets. source ↗
Training Data Providers such as banks, hospitals, and government agencies would make data accessible through the framework. source ↗
The architecture rests on four pillars: confidential clean rooms, differential privacy, electronic contracts, and zero-trust security. source ↗
The framework targets DPDP Act compliance, was drafted with think tank iSPIRT, and extends the existing DEPA used by Account Aggregator. source ↗

Practical applications

Teams building India-specific models should inventory which training-data gaps could be filled by Training Data Providers like banks, hospitals, and government agencies if the framework is adopted.
ML engineers should prototype training runs inside confidential clean rooms with differential privacy now to learn the accuracy and infrastructure trade-offs before compliance demands it.
Data-holding organisations should draft the electronic-contract terms — permitted uses, security and protection requirements — under which they would participate as providers.
AI companies with India operations should engage with MeitY while the framework is still a proposal, since the final consent and zero-trust requirements will shape their data pipelines.

Context

India already runs DEPA for consent-based data sharing in finance through the Account Aggregator system, and this proposal extends that architecture to AI training data. The DPDP Act is India's data-protection law, so large-scale dataset sharing must comply with it. Techniques like differential privacy and clean rooms aim to let models learn from data without exposing individual records.

What to watch

Watch whether MeitY finalises the guidelines and names pilot deployments or the first Training Data Providers.
Any conflict between the framework's consent mechanics and DPDP Act enforcement would signal a redesign.

Related briefs

Editorial score 3.4 / 5 · significance 3.5 · novelty 4.0 · edge 3.0 · perspective 3.0

Desks: Policy & Society · Business

Topics: Governance & policy · AI research

Evidence basis: Reviewed from the article's full text

This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.

MeitY Floats New Privacy-Led Framework To Unlock Data For Training AI Models

Original reporting by Inc42 — AI — Gaurav Bagur · published Oct 30, 2026, 3:22 PM EDT

Indian startups and researchers could gain privacy-compliant access to large domestic datasets from banks, hospitals, and government agencies for training AI models.

Key points

MeitY has proposed DEPA for AI training, a consent-based framework for sharing data to train AI models in India. source ↗

Organisations could collaborate on model training without directly transferring the underlying datasets. source ↗

Training Data Providers such as banks, hospitals, and government agencies would make data accessible through the framework. source ↗

The architecture rests on four pillars: confidential clean rooms, differential privacy, electronic contracts, and zero-trust security. source ↗

The framework targets DPDP Act compliance, was drafted with think tank iSPIRT, and extends the existing DEPA used by Account Aggregator. source ↗

Practical applications

Teams building India-specific models should inventory which training-data gaps could be filled by Training Data Providers like banks, hospitals, and government agencies if the framework is adopted.

ML engineers should prototype training runs inside confidential clean rooms with differential privacy now to learn the accuracy and infrastructure trade-offs before compliance demands it.

Data-holding organisations should draft the electronic-contract terms — permitted uses, security and protection requirements — under which they would participate as providers.

AI companies with India operations should engage with MeitY while the framework is still a proposal, since the final consent and zero-trust requirements will shape their data pipelines.

Context