6 Best Practices For Securing the CI/CD Pipeline to Prevent Data Leakage

Grace Lau -
cicd pipeline security
Illustration: © IoT For All

Development teams are under constant pressure to meet deadlines and release products as quickly and profitably as possible. This has been enabled by continuous integration (CI) and continuous delivery/deployment (CD).  Product and profit are not the only concerns when working in development. The CI/CD pipeline, like any system, can be vulnerable to data leakage, and security steps must be taken to prevent this wherever possible.

We will take a look at the CI/CD pipeline, what makes it vulnerable to data leakage, and what practices can be implemented to ensure against this.

What Is the CI/CD pipeline?

To understand how to secure the CI/CD pipeline effectively, we must first know what this is. Essentially, the CI/CD pipeline is the cyclical process for development that combines continuous integration, delivery, and deployment. 

Continuous integration mainly focuses on the earliest stages of the pipeline, when code is written and undergoes initial testing. It often involves multiple developers working simultaneously, making frequent small changes to the code, allowing for low-risk experimentation.

On the other hand, continuous delivery makes up the later stages of the pipeline. This involves the thorough testing of the build, leading up to its validation and deployment. Again, incremental changes are made to allow for the easy roll back to earlier builds if errors are discovered.

The main difference between continuous delivery and continuous deployment is that deployment automatically sends each validated build to production. In contrast, delivery readies the validated build for manual authorization and deployment. 

This makes continuous deployment faster, but it can also allow undetected vulnerabilities to slip through to a production build undiscovered.

There are many benefits to the CI/CD pipeline. The greatest of which is that it’s incredibly efficient- given that it’s primarily automated through systems such as Google RPA. 

However, this level of automation and the speed with which the pipeline develops can leave it vulnerable to data leakage, which is why it’s crucial to ensure the CI/CD pipeline is secure.

Why Is the CI/CD Pipeline At Risk?

The CI/CD pipeline is at risk for several reasons:


The fact that so much of the CI/CD pipeline relies on automation means that sophisticated attack methods and malware can often go undetected until it’s too late, as gaps in the systems’ defenses are not continuously monitored.

Limited Testing

Unfortunately, so many different developers are contributing to the CI/CD pipeline can mean that security gaps can appear. Poor coordination of resources can lead to limited or forgotten test cases, meaning errors can slip through the testing process and be exploited further down the line.

Cloud-Hosted Services

Wherever cloud-hosted or open-source services are employed, vulnerabilities can arise. Not having sufficient cyber security measures in place in something as simple as a system or protocol for using your virtual phone number can leave the whole platform open to attack from security threats.

Now that we’ve seen how the CI/CD pipeline can be put at risk from security threats, we can examine its protection against malicious attacks. Here are six best practices for securing the pipeline to prevent data leakage.

Threat Modeling

A threat modeling exercise is a great place to start when looking at any cybersecurity. After all, how are we supposed to defend against threats if we don’t know what they are or where they may strike? Threat modeling allows us to identify relevant threats to the pipeline and vulnerabilities where these may gain access.

Threat modeling can be used at every stage of the development life cycle. Diagrams and flowcharts can be helpful here, allowing you to visualize every step in the pipeline. You’ll need to construct an overview of your CI/CD pipeline and then decompose every element within this to gain a detailed understanding of the mechanics at play. 

Once you’ve laid out the map of your pipeline and are aware of the different stages within it, you can begin to identify where it may be under threat. Software application testing and similar methods can be used here to identify vulnerabilities. 

A good idea when threat modeling is to rate and rank the threats you’ve identified, allowing you to prioritize which vulnerabilities need addressing first.

Audit and Monitor the Pipeline

One of the best ways to minimize the risks to your CI/CD pipeline is to be aware of what it’s composed of and what is traveling through it. As mentioned above, threat modeling is a great place to start with this.

But auditing the pipeline is a process that can’t simply be performed once or twice. It needs to be happening almost constantly. The pipeline can be ever-changing, and unfortunately, the more complex it becomes, the more vulnerabilities can arise within it. A complex pipeline may have vulnerabilities that are easily missed, and an attack upon them can go entirely unnoticed. 

Keeping detailed “maps” of your pipeline environments and the tools used within them will allow you to keep track of what kind of information is processed and stored at each stage. This will help you determine which pipeline areas need the most careful monitoring for threats.

System logs are invaluable when auditing your pipeline and should be looked at alongside repository, build, and deployment logs to monitor activity along the pipeline. Once you know what’s happening at each pipeline stage, you can add relevant security measures to keep it secure.

Stay Updated

Security vulnerabilities in the software and apps your pipeline uses can leave it open to cyber-attacks and data leakage. Developers will constantly update their products to identify their weaknesses and release patches to fix them. 

Keeping up to date with regular patches for everything from word processing software to outbound calling solutions is therefore essential to the security of your CI/CD pipeline.

Automatic updates are a great way to save time and effort for keeping your software up to date. But, it’s also important to watch out for unexpected, critical updates that may be released outside of the standard development roadmap.

However, you must be careful not to apply all updates which appear for download blindly. The updates themselves can be points of vulnerability that can be exploited. This is what happened in the SolarWinds hack in 2020. Checking for vulnerabilities on the CVE site, especially before applying major updates, can save a lot of hassle.

Software composition analysis (SCA) tools can analyze third-party apps and software to look for security alerts and monitor open source code that could be corrupted. 

Similarly, when choosing third-party software and apps to utilize, let’s say a warehouse management software, look for indicators such as trust badges, ISO certification, and product reviews to make sure you’re downloading legitimately secure products.

Monitor Access

It may seem too obvious to mention, but one of the best ways to prevent data leakage is to keep a close eye on who is using your systems and why. There may be a large number of active users accessing your pipeline, often remotely. 

You should be wary about which users are given what privileges. Always abide by the principle of least privilege. Simply put, only allow users access to the absolute minimum they need to fulfill their responsibilities. Keeping track of who requires what privileges can be aided with workforce management tools. For businesses that rely upon on-call services or calls in general, the cloud PBX system must be protected at all costs.

Create user roles and grant privileges discriminately. Allowing developers access only to the tools they need prevents unnecessary risk. It reduces the chance of human error from untrained developers accidentally accessing software or code that they’re not fully trained in the application of.

As with the pipeline itself, permissions and access should be audited continuously. Any stale users and applications should be removed from the system to reduce the number of entry points that could be exploited. It would help if you also were mindful of any suspicious behavior, such as users attempting to access systems they don’t have permission to use.  

As well as monitoring users and their permissions, it would help if you also were mindful of automated requests and processes. Automation tools can save an awful lot of time, but they can also present points of vulnerability that can be exploited. If you see abnormal automated pull requests, these should be investigated thoroughly, as they can be used to execute malware.

Utilize Cybersecurity Tools

Utilizing cybersecurity tools is an essential but all too often overlooked aspect of securing the CI/CD pipeline. Making sure any developer with access to the pipeline runs rigorous anti-spyware and malware programs seems too obvious to mention. Still, we should make sure we’ve covered the basics, too!

As the old saying goes, “prevention is better than the cure.” You can take many steps to mitigate security risks, which are all the more essential when your network is spread across multiple internet-enabled devices. 

Alongside the bread and butter of internet security protocols such as firewalls, and data encryption, more niche measures can be implemented to secure your pipeline against data leakage:

  • Password Managers. Using different passwords for different software along the pipeline can prevent leakages at one point from spreading along the whole pipeline. A password manager can be used to secure the various passwords behind military-grade encryption.
  • One-time passwords. Areas of the pipeline that require even heftier security can be locked behind one-time passwords generated as and when access is required.
  • IDE plugins. These can guard against data leakage by providing fixes during the actual writing of code by detecting coding errors and highlighting vulnerabilities 

Confidential Information

The simplest way to avoid sensitive data leaking from your CI/CD pipeline is to not put it there in the first place. There are some instances where this will not be possible, but every piece of unnecessary information added to the pipeline is another piece that could be needlessly leaked.

Personally identifiable information gathered through website personalization tools, for example, should not be stored in accessible repositories when it’s not necessary to do so. Or, if it is necessary to do so, make sure they’re stored in an account or resource that has the least privileges required to perform the task for which they’re needed. 

It’s also worth remembering that some methods of communication are more susceptible to infiltration than others. Email inboxes are notoriously easy to hack, so sharing sensitive information via email between developers should be discouraged where possible. 

Try looking into alternative solutions, such as a fixed VoIP phone. Where it’s unavoidable to communicate electronically, make sure the appropriate security measures are in place.

Now Your Data Is Secure

There we go. The pipeline is secure. Or as secure as it can be. These six best practices are by no means all that is required to keep your CI/CD pipeline secure, but they’re an excellent place to start. 

Each CI/CD pipeline will need unique security measures to keep it secure against the threats it may face. Be aware of what composes the pipeline and where it’s most vulnerable, and you will be well prepared.

Implementing these practices is more manageable when everyone with access to the pipeline is working to the same guidelines. Therefore, it’s essential to ensure that you have clearly defined development processes that all your developers and testers are aware of and are working together to uphold. 

And finally, don’t forget that as your pipeline evolves, so will the threats facing it. Be constantly vigilant and mindful of what new measures you may need to take in the future.

Grace Lau

Guest Writer
Guest Writer
Guest writers are IoT experts and enthusiasts interested in sharing their insights with the IoT industry through IoT For All.
Guest writers are IoT experts and enthusiasts interested in sharing their insights with the IoT industry through IoT For All.