Guest Post by C.J. Radford, VP of Cloud for Vormetric

Big Data initiatives are heating up. From financial services and government to healthcare, retail and manufacturing, organizations across most verticals are investing in Big Data to improve the quality and speed of decision making as well as enable better planning, forecasting, marketing and customer service. It’s clear to virtually everyone that Big Data represents a tremendous opportunity for organizations to increase both their productivity and financial performance.

According to WiPro, the leading regions taking on Big Data implementations are North America, Europe and Asia. To date, organizations in North America have amassed over 3,500 petabytes (PBs) of Big Data, organizations in Europe over 2,000 PBs, and organizations in Asia over 800 PBs. And we are still in the early days of Big Data – last year was all about investigation and this year is about execution; given this, it’s widely expected that the global stockpile of data used for Big Data will continue to grow exponentially.

Despite all the goodness that can stem from Big Data, one has to consider the risks as well. Big Data confers enormous competitive advantage to organizations able to quickly analyze vast data sets and turn it into business value, yet it can also put sensitive data at risk of a breach or violating privacy and compliance requirements. Big Data security is fast becoming a front-burner issue for organizations of all sizes. Why? Because Big Data without security = Big Risk.

The fact is, today’s cyber attacks are getting more sophisticated and attackers are changing their tactics in real time to get access to sensitive data in organizations around the globe. The barbarians have already breached your perimeter defenses and are inside the gates. For these advanced threat actors, Big Data represents an opportunity to steal an organization’s most sensitive business data, intellectual property and trade secrets for significant economic gain.

One approach used by these malicious actors to steal valuable data is by way of an Advanced Persistent Threat (APT). APTs are network attacks in which an unauthorized actor gains access to information by slipping in “under the radar” somehow. (Yes, legacy approaches like perimeter security are failing.) These attackers typically reside inside the firewall undetected for long periods of time (an average of 243 days, according to Mandiant’s most recent Threat Landscape Report), slowly gaining access to and stealing sensitive data.

Given that advanced attackers are already using APTs to target the most sensitive data within organizations, it’s only a matter of time before attackers will start targeting Big Data implementations. Since data is the new currency, it just makes sense for attackers to go after Big Data implementations because that’s where big value is.
So, what does all this mean for today’s business and security professionals? It means that when implementing Big Data, they need to take a holistic approach and ensure the organization can benefit from the results of Big Data in a manner that doesn’t negatively affect the risk posture of the organization.
The best way to mitigate risk of a Big Data breach is by reducing the attack surface, and taking a data-centric approach to securing Big Data implementations. These are the key steps:

Lock down sensitive data no matter the location.

The concept is simple; ensure your data is locked down regardless of whether it’s in your own data center or hosted in the cloud. This means you should use advanced file-level encryption for structured and unstructured data with integrated key management. If you’re relying upon a cloud service provider (CSP) and consuming Big Data as a service, it’s critical to ensure that your CSP is taking the necessary precautions to lock down sensitive data. If your cloud provider doesn’t have the capabilities in place or feels data security is your responsibility, ensure your encryption and key management solution is architecturally flexible in order to accommodate protecting data both on-premise and in the cloud.

Manage access through strong polices.

Access to Big Data should only be granted to those authorized end users and business processes that absolutely need to view it. If the data is particularly sensitive, it is a business imperative to have strong polices in place to tightly govern access. Fine-grained access control is essential, including things like the ability to block access by even IT system administrators (they may have the need to do things like back up the data, but they don’t need full access to that data as part of their jobs). Blocking access to data by IT system administrators becomes even more crucial when the data is located in the cloud and is not under an organization’s direct control.

Ensure ongoing visibility into user access to the data and IT processes.

Security Intelligence is a “must have” when defending against APTs and other security threats. The intelligence gained can support what actions to take in order to safeguard and protect what matters – an organization’s sensitive data. End-user and IT processes that access Big Data should be logged and reported to the organization on a regular basis. And this level of visibility must occur whether your Big Data implementation is within your own infrastructure or in the cloud.

To effectively manage that risk, the bottom line is that you need to lock down your sensitive data, manage access to it through policy, and ensure ongoing visibility into both user and IT processes that access your sensitive data. Big Data is a tremendous opportunity for organizations like yours to reap big benefits, as long as you proactively manage the business risks.

You can follow C.J. Radford on Twitter @CJRad.