The global AI race is heating up, but India is carving its own path. A new framework called DEPA aims to solve the biggest puzzle in AI: how to use valuable data for innovation while giving citizens unprecedented control over their privacy.
Demystifying DEPA: India's New Data Paradigm
At its core,
the Data Empowerment and Protection Architecture (DEPA) is a visionary framework designed to revolutionize how data is shared and used in India. Spearheaded by the Ministry of Electronics and Information Technology (MeitY) with support from the iSPIRT Foundation, DEPA is the next evolution of the India Stack—the digital public infrastructure that gave us UPI and Aadhaar. Instead of companies holding user data in proprietary silos, DEPA creates a system where individuals are the true owners of their data. It establishes a secure, consent-based method for people to share their information for specific purposes, from getting a loan to receiving medical advice. This shift is not just a policy update; it's a fundamental change in data governance, moving from a model of data collection to one of data empowerment.
The Power of Granular Consent
The magic of DEPA lies in its consent mechanism. Today, we often agree to lengthy terms and conditions without fully understanding what data we are giving away. DEPA replaces this broken system with 'Consent Managers'—neutral third-party platforms that allow users to grant and manage data-sharing permissions with granular detail. Through a simple digital interface, you can decide exactly what data to share (e.g., bank statements for a three-month period), with whom (e.g., a specific lender), for what purpose (e.g., a home loan application), and for how long. This consent is machine-readable, auditable, and easily revocable, creating a transparent trail that holds data users accountable and puts the individual firmly in control.
Fueling AI Without Sacrificing Privacy
The latest extension of this framework, 'DEPA for AI Training', directly addresses the AI industry's insatiable hunger for data. Training powerful AI models requires vast and diverse datasets, which are often locked away in sensitive sectors like healthcare and finance. DEPA for AI Training solves this 'Data Deadlock' with an ingenious principle articulated by India's Principal Scientific Adviser, Prof. Ajay Kumar Sood: 'the model travels to the data, rather than the data travelling to the model'. Instead of exporting raw, sensitive data, AI developers will run their training algorithms within secure, controlled computing environments hosted by the data custodians themselves. This process uses advanced privacy-enhancing technologies like differential privacy and confidential computing to ensure developers never see the raw data, only the insights needed to improve their models.
From Theory to Real-World Impact
The implications of this are immense. In finance, DEPA could help extend formal credit to 400 million underserved individuals and small businesses by allowing lenders to assess creditworthiness based on a wider, yet consented, range of data. In healthcare, researchers could train diagnostic AI on anonymized data from multiple hospitals to detect diseases earlier, without compromising patient confidentiality. The framework aims to unlock an estimated $500 billion in AIndriven value for India's GDP by 2030 by enabling this kind of responsible innovation. It levels the playing field, allowing startups to compete with large tech companies by giving them governed access to high-quality datasets that were previously out of reach.
The Road Ahead : A Blueprint for the World?
While DEPA for AI Training is currently a techno legal framework with pilot programs planned, its ambition is global. It presents a clear alternative to the data extraction models of the West and the state controlled approach elsewhere, positioning India as a leader in responsible AI governance. This non aligned, technical first approach could serve as a blueprint for other developing nations in the Global South, allowing them to build sovereign AI capabilities without surrendering their citizens' data. The success of DEPA will depend on wide scale adoption, robust technical standards, and building trust across the ecosystem. However, it signals a strategic move to build an AI future that is not only innovative but also equitable and secure.
