The 'Shared Responsibility' Misnomer: Why the Cloud Continues to Confound

Organizations of all sizes and industries are embracing cloud solutions for the ease of standing up software, infrastructure, and platform resources. However, security fears linger, and rightly so. In the periphery of these IT investment decisions are justifiable concerns, including data leakage and misconfigurations, which impede a broader and deeper embrace of cloud capabilities.

Cloud vendors describe the security management of cloud offerings as a "shared responsibility model," and the technical community has quickly adopted this vernacular. It conveys an equitable arrangement: The cloud services provider (CSP) will handle half of the responsibility, and the customer will handle half. All the customer needs to do is understand and apply their part of the security responsibility, and they should be good to go. Thus, the breaches we read about are presumably attributable to an organization being woefully negligent in learning and applying its half of the responsibility pie.

Yet, not typically understood is the scope and complexity of the customer's share of the responsibility. At its core, the shared responsibility model means the CSP is responsible for securing the infrastructure of the cloud, and the customer is responsible for security in the cloud — the data itself and the controls and settings that protect that data. What is commonly misunderstood is the vast nature of the customer's world of responsibility and the complexity within the controls environment that confounds attempts at ease of management.

This is not intended to disparage cloud computing. The cloud has enabled companies to scale faster and more efficiently than they could otherwise, as well as access capabilities they could not easily develop and maintain themselves, all in a cost-effective manner. Rather, it is to put the cloud challenge into perspective and help organizations get a stronger understanding of what they must consider when planning a cloud security strategy.

Services, Controls, Settings, and Complexity Spill Far Outside of Buckets
Today, organizations have more cloud platforms from more vendors than ever before. In Flexera's 2020 State of the Cloud Report, the number of companies spending between $2.4 million and $12 million on the cloud increased 20% in 2019, and 93% have a multicloud strategy. Notably, 81% cited security in the cloud as their top cloud challenge.

When we inspect the challenge even more closely, the complexity within each cloud platform becomes more evident. CSPs offer more products and services within their platforms every year, on top of already-rich offerings. This is a good thing for customers: They have more capabilities within the same vendor's platform. However, virtually every product and service have security controls and settings that must be learned, applied, and rechecked regularly.

In addition, each CSP is different, and the security tools designed to monitor and secure them rarely cross platforms. Thus, the shared responsibility model for a company with five platforms requires reading, understanding, and diligently applying sometimes esoteric knowledge from extensive documentation across five vendors.

There is ripe opportunity for error within just the cloud environment, not to mention the other cybersecurity tasks related to on-premises, private cloud, and legacy IT assets. For the small to midsize business, skills gaps are often the issue; for the large enterprise, the vast, diverse, dispersed, and constantly changing IT estate can be challenging.

Understanding (and Preparing for) the Challenge
In our experience as incident responders, threat actors are scanning for every doorway left open. While user misconfigurations of S3 bucket asset controls seem to get the focus of "shared responsibility errors," attackers are looking for any vulnerability as a foothold to gain entry. We have seen attackers obtain access to internal cloud resources through a number of innovative means, such as taking advantage of network appliance misconfigurations and through the discovery of developer API keys in GitHub repositories (either compromised or publicly accessible).

While it may seem that organizations are negligent when a cloud breach happens, it is important to note that not only are threat actors numerous, watchful, and sophisticated, they are also creative and will leverage many different avenues in their attempts to gain access.  

Before diving into, or diving deeper, into the cloud, ensuring your organization is secure requires understanding the breadth of the challenge, assessing skills and technology gaps, and then devising a security strategy that encompasses the people (focused expertise), processes, and technology (including cloud-centric security tools) to meet it. While companies could conceivably throw endless dollars at the security challenge, a full embrace of the cloud without a security strategy can have significant costs of its own through potential breach costs and brand damage. That strategy can include on-site staff, managed security services, or a combination of the two.

Above all, companies diving deep in the cloud should be fully aware going in that, while they may be sharing the responsibility, the task on the customer end is quite complex and expansive. It may only take a matter of hours to spin up a new cloud capability. However, the skills to defend the data in the cloud is another challenge that must be met, and quickly, before threat actors sniff out an opportunity to act.