Data Protection In The Cloud
2015 ushered in the start of a data economy. As organizations amass more detailed consumer profiles they have begun realizing that data could equal or surpass the value of the products and services they sell, especially in the Internet of Things era with its constant and very personal streams of data. Data breaches such as the Office of Personal Management and toymaker, VTech are indicative of increasing hactivist interest in more personal data and also of the growing value of that data.
At the same time the concept of cloud is changing. In our hyper connected era traditional backend clouds where the bulk of data processing takes place have been superseded by waves of cloud migration that are closer to where the data transaction is occurring. This allows for real-time data exchanges. Additionally, the lines between SaaS, PaaS and IaaS are becoming blurred with hybrid models such as SaaS built upon PaaS. With the confluence of a data economy, blurring of cloud models, and far more egregious data breaches I have outlined principles that Information Security Practitioners may want to consider as we move into 2016.
1. Bake standard data security profiles into a cloud brokerage platform that can be applied on as needed consumption basis. This will more easily allow IT and InfoSec to keep pace with new instantiations by the business across the cloud-extended data center.
2. Place increasing importance on federated identity schemes with individuals having multiple devices across different cloud services.
3. Build a data brokerage to help calculate the value of data. It’s the most effective way for business users to learn the value of the data they create, collect or handle.
Protect data according to the following domains:
- State data classification in business consumable terms if you want business users to own up to protecting data according to its business risk.
- Leverage machine learning for dynamic data classification as data changes value over the course of its lifecycle.
- Where possible digitally tag or watermark data that is transacted, stored or processed with a cloud provider. This minimizes confusion around data ownership and entitlement rights.
Data Protection and Lifecycle Management
- Ensure policy management extends to access management at the various admin layers for the cloud provider as well as for the elements of the cloud stack you as an organization have control over.
- Enable data owners to specify what actions users can take– read, write, copy, modify.
- Ensure that data lifecycle management – creation, modification, retention, destruction is built into your policies.
- Set encryption settings – key strength and key management parameters based on data sensitivity.
- Continuously log all actions based on the context of who, what when and where.
By Evelyn de Souza