This is a weekly post where we address questions of interest to the Application Information Security Community. To that end, we’d love to hear your questions! Please Tweet us with the hashtag #ThinkAppSec or leave us a comment below and we’ll pick one or two questions from that list.

How to assess the security of a cloud service provider

This is a question that I hear a lot. And it’s been asked since the early days of ASP (Application Service Providers). While it’s a simple question, the answer is fairly complex. Below we outline some of the high-level points organizations should consider when looking at their CSP’s application security profile.

How do I know my Cloud Service Provider’s (CSP) Applications are secure?

1. Know What You Mean by “Secure”

The most important step a company can take is to define what they mean by “secure.” If you don’t know what you’re aiming for, it’s going to be hard to know if you’ve hit the target. A rule of thumb that I’ve found useful in my consulting past is to review the security requirements for the application or service that was being used in-house and ensuring that the same requirements are applied to that app or service in the cloud. In other words, if your organization required that sensitive data be stored encrypted when it was on-prem, that data should probably be encrypted in the cloud – and arguably there could be even more reason to have it encrypted at the cloud provider.

When defining application security requirements for the cloud, don’t forget physical and personnel protections and restrictions. Your data center may have 24/7 camera surveillance, 3 layers of back-up energy supply and biometric authentication for entry – what does the cloud provider have? Hiring practices vary by company. Do you require a full background check of employees in the data center? If so, do you need that same level of personnel assessment from your CSP for the people managing your application?

Another point to consider is geographic location. Countries have different approaches to data privacy and access. When your application or service is housed in your data centers, you control where the data resides. A CSP may have optimized operations all over the globe. If you have organizational restrictions on countries where your data is stored, you’ll want to make sure the CSP can keep your data in the right geo-location.

2. Understand Audit Options

Some organizations feel comfortable if their CSP can provide any kind of assurance report, like a PCI DSS RoC (Payment Card Industry Data Security Standard Report on Compliance), ISO/IEC 27001 certification, or a SSAE 16 SOC2 (was SAS 70 Type II) report. While these reports are useful, they often don’t tell the whole risk story for a CSP or the application you are using in their Cloud.

Another approach is to include a “right to audit” clause in the contract with the CSP which enables you, as the customer, to perform some kind of audit tasks. What those tasks are will depend on what is defined in the contract. Some customers go so far as to require quarterly on-site visits for physical assessment and regular vulnerability scans and penetration tests of the CSP’s cloud/application. While this approach is comprehensive, getting that level of audit capability isn’t easy and many CSPs won’t agree to those audit terms. Not to mention that you, as the customer, probably have to pay the CSP more for the audit access, and have to allocate resources to perform the audits, review the findings and work with the CSP to ensure that critical issues are remediated.

3. Consider CSA STAR (Cloud Security Alliance Security Trust & Assurance Registry)

Once you know your requirements and have any required assurance reports – rather than going to a full audit of the CSP yourself, your organization could check out the CSP’s in the CSA STAR.

“The CSA Security, Trust & Assurance Registry (STAR) is a publicly accessible registry that documents the security controls provided by various cloud computing offerings, thereby helping users assess the security of cloud providers they currently use or are considering contracting with.”

There are three layers to the CSA STAR Open Certification Framework:

Both the CAI and the CCM cover much more than application security and can be used by an organization to assess the CSP’s overall security stance as well as to nail down specifics about what they’re doing to provide security and protection for their applications and the data stored or used by those apps. Current STAR Certification means that a provider has been assessed against CCM v1.4 and ISO/IEC 27001:2005. In March 2014 providers can be assessed against either CCM v1.4 or CCM v3.

The first control domain of the CCM focuses on Application and Interface Security and provides control specifications for that domain including application security and data integrity. But many of the other domains touch on application security – the CCM has these identified in the Architectural Reference column so organizations can find the domains and control specifications quickly.

Organizations can review the level that a provider has completed (self-assessment, certification, continuous) in the Registry as a precursor to selecting a provider and to get some answers about what they’re already doing for application security. The CCM and STAR Registry may not answer every question you have about a CSP’s application security, but it is a great place to start. And a good way to get your thinking going about what else you want to know from the vendor, map your requirements to what is in the CAI or the CCM to see where there are gaps and ask the vendor about those during the assessment or contract negotiation process.

Clearly, there are many more steps involved in doing a complete assessment of a CSP’s application security. But hopefully the points above will serve as a good starting point to get your thinking organized and strategy for assessment created.

What are you doing to assess and monitor your CSP’s application security? What has worked for you? And what hasn’t? Please let us know in the comments below.

Other Application Security Questions Answered

What is the importance of software security in supply chain management?

Who Should be Responsible for Application Security Testing?

Can “generated code” be tested?

How do we secure application vulnerabilities and code development, particularly for mobile and social applications that are built by business units or reside on the cloud?

As a CISO, how can I control my organization’s testing methodologies, change management and deployment processes, without compromising on quality and project timelines?
How Can I Secure Apps in the Cloud?

Will the legal landscape change if software vendors can be sued without damages or loss being proven?
The Legal Landscape: Can vendors be sued without damages? What the heck is PII?

What is PII – How much can the definition expand?
Mobile Apps: Which are More Secure Android or iOS?

Does IoT (Internet of Things) “change everything” for Application Security?

What is the difference between PCI DSS and PA DSS?

How can we foster cooperation to help our Development and Security Teams work together?

How do I know my Cloud Service Provider (CSP) Applications are secure?

What can I do to help eradicate SQLi or at least reduce the incidence of SQLi vulns in our production applications?

Submit your questions via Twitter using #ThinkAppSec

More from Application Security

PixPirate: The Brazilian financial malware you can’t see

10 min read - Malicious software always aims to stay hidden, making itself invisible so the victims can’t detect it. The constantly mutating PixPirate malware has taken that strategy to a new extreme. PixPirate is a sophisticated financial remote access trojan (RAT) malware that heavily utilizes anti-research techniques. This malware’s infection vector is based on two malicious apps: a downloader and a droppee. Operating together, these two apps communicate with each other to execute the fraud. So far, IBM Trusteer researchers have observed this…

From federation to fabric: IAM’s evolution

15 min read - In the modern day, we’ve come to expect that our various applications can share our identity information with one another. Most of our core systems federate seamlessly and bi-directionally. This means that you can quite easily register and log in to a given service with the user account from another service or even invert that process (technically possible, not always advisable). But what is the next step in our evolution towards greater interoperability between our applications, services and systems?Identity and…

Audio-jacking: Using generative AI to distort live audio transactions

7 min read - The rise of generative AI, including text-to-image, text-to-speech and large language models (LLMs), has significantly changed our work and personal lives. While these advancements offer many benefits, they have also presented new challenges and risks. Specifically, there has been an increase in threat actors who attempt to exploit large language models to create phishing emails and use generative AI, like fake voices, to scam people. We recently published research showcasing how adversaries could hypnotize LLMs to serve nefarious purposes simply…

Topic updates

Get email updates and stay ahead of the latest threats to the security landscape, thought leadership and research.
Subscribe today