The value and vulnerability of centralized data storage

News:  Data-sharing infrastructure projects in India are accelerating.

The most visible developments in Data-sharing infrastructure projects are seen in the financial sector. For instance, the recently-launched account aggregator framework made it possible for over 300 million users to share their data with different financial entities across the credit ecosystem.

Similar activity is also underway in many other sectors, notably the health sector.

In this situation, there is a debate on how to build robust digital infrastructure. This article analyses the challenges and advantages of the proposed options to build digital infrastructure in India.

What are the suggestions put forwarded to build robust digital infrastructure in India?

Data to be stored in the central repository: data should be extracted from the silos (repositories) in which they currently reside and aggregated into a central repository.

Federated model: data should be simply left where they are but interconnected, so that the data, at the directive of the user, can be transferred from one entity to another.

What are the challenges and advantages of the proposed options?

Data to be stored in a central repository

Advantage: A central repository makes data management easier by allowing individuals greater control over their data without the need to rely on data collectors to provide them access to their own data.

Disadvantage:  A security breach in the central repository would place the sensitive personal data of the entire population at risk.

Federated model

Advantage: it ensures that all this data does not end up in a single archive. If data can remain where it currently is and be transferred on request, this model offers the same benefits with significantly less risk.

Disadvantage

Data users want their personal data to be available indefinitely so that they can use it many years and even decades into the future.

Data collectors, on the other hand, have no incentive to store the data any longer than necessary, particularly since storage comes at a cost and their liability for a data breach is on the rise.

Further, these entities could go bankrupt or change the focus of their business, and the data under their care could be lost forever.

What are the solutions proposed to tackle this issue?

The government should establish a centralized data repository and offer it as a public good for the benefit of all its citizens, along the lines of Digi Locker.

Data that is currently stored in different databases throughout the federated data architecture could be backed up by this repository. It will ensure that even if any one of the service providers goes down, the data is not lost.

What are the issues in the proposed solution?

It would be a bad idea because this is not a job for the state. The core function of government is to govern.

The government should only be involved in prescribing the data sharing protocols, establishing the regulations that will govern participants in the ecosystem, and ensuring compliance.

Source: This post is based on the article “The value and vulnerability of centralized data storage” published in Livemint on 29^th Dec 2021.