Data Context is the Missing Ingredient for Security Teams
Why are we still struggling with remediation and alert fatigue? In every cybersecurity domain, as we get better at identifying vulnerabilities, and add new automation tools, security teams still face the same challenge - what do we remediate first? What poses the greatest risk to the business?
Of course, the capabilities of cyber solutions have grown. We have more information about breaches and potential risk than ever. If in the past, an EDR could tell you which endpoint has been compromised, today an XDR can tell you which servers and applications have been compromised. It’s a deeper level of analysis. But prioritizing what to focus on first is still a challenge. You might have more information, but it’s not always clear what the biggest risk to the business is.
The same can be said for SIEMs and SOAR solutions. If in the past we received alerts and made decisions based on log and event data from the SIEM, now we can factor in threat intelligence and third party sources to better understand compromises and vulnerabilities. But again, when it comes to what to remediate to best protect your specific business these tools aren’t able to prioritize.
The deeper level of analysis we’ve been conducting for the last 5-10 years is still missing what’s needed to make effective remediation recommendations - context about the data at risk. We get all these alerts, and while we might know which endpoints and applications are affected, we’re blind when it comes to the data. That ‘severe’ endpoint vulnerability your team is frantically patching? It might not contain any sensitive data that could affect the business. Meanwhile, the reverse might be true - that less severe vulnerability at the bottom of your to-do list might affect data stores with customer info or source code.
AWS CISO Stephen Schmidt, showing data as the core layer of defense at this years AWS Reinforce
This is the answer to the question ‘why is prioritization still a problem?” - the data. We can’t really prioritize anything properly until we know what data we’re defending. After all, the whole point of exploiting a vulnerability is usually to get to the data.
Now let’s imagine a different scenario. Instead of getting your usual alerts and then trying to prioritize, you get messages that read like this:
‘Severe Data Vulnerability: Company source code has been found in the following unsecured data store:____. This vulnerability can be remediated by taking the following steps: ___’.
You get the context of what’s at-risk, why it’s important, and how to remediate it. That’s data centric security.
Why Data Centric Security is Crucial for Cloud First Companies
Data centric security wasn’t always critical. When everything was stored on the corporate data center, it was enough to just defend the perimeter, and you knew the data was protected. You also knew where all your data was - literally in the room next door. Sure, there were risks around information kept on local devices, but there wasn’t a concern that someone would accidentally save 100 GB of information to their device.
The cloud and data democratization changed all that. Now, besides not having a traditional perimeter, there’s the added issue of data sprawl. Data is moved, duplicated, and changed at previously unimaginable scales. And even when data is secured properly, with the proper security posture, that security posture doesn’t come with when the data is moved. Legacy security tools built for the on-prem era can’t provide the level of security context needed by organizations with petabytes of cloud data.
Data Security Posture Management
This data context is the promise of data security posture management solutions. Recently recognized in Gartner’s Hype Cycle for Data Security Report as an ‘On the Rise’ category, DSPM gets to the core of the context issue. DSPM solutions attack the problem by first identifying all data an organization has in the cloud. This step often leads to the discovery of data stores that security teams didn’t even know existed. Following this, the next stage is classification, where the types of data labeled - this could be PII, PCI, company secrets, source code, etc. Any sensitive data found to have an insufficient security posture is passed to the relevant teams for remediation. Finally, the cloud environment must be continuously assessed for future data vulnerabilities which are again forwarded to the relevant teams with remediation suggestions in real time.
In a clear example of the benefits offered by DSPM, Sentra has identified source code in open S3 buckets of a major ecommerce company. By leveraging machine learning and smart metadata scanning, Sentra quickly identified the valuable nature of the exposed asset and ensured it was quickly remediated.
If you’re interested in learning more about DSPM or Sentra specifically, request a demo here.
