How to Mitigate Low-Code Security Risks

Content By Devops .com

Gartner predicts that by the end of 2025, over 65% of development projects will usee low-code builders. The field of low-code continues to expand. But what security implications does low-code introduce?

Low-code refers to tools that enable application construction using visual programming models. Adopting drag-and-drop components instead of traditional code, no-code and low-code platforms enables non-technical folks to construct their own workflows without as much help from IT. Yet, handing power to citizen developers with less security training can be risky. Plus, low-code platforms may hold compromised propriety libraries or leverage APIs that may unknowingly expose sensitive data to the outside world. There’s also the possibility that low-code could increase shadow IT if not governed well.

Low-code opens the door to more producers. But, does it open the backdoor, too? I recently met with Chris Wysopal, CTO and cofounder of Veracode, to discuss the growing security risk posed by low-code applications. According to Chris, hardening low-code environments involves a combination of compartmentalization, dynamic and runtime techniques, and adopting the “least privilege as possible” rule across the board. Below, we’ll expand on potential low-code risks and see what other methods Wysopal recommends organizations employ to secure the emerging low-code paradigm.

Low-Code Security Concerns

Verizon’s 2020 Data Breach Investigations Report (DBIR) found 43% of all breaches are connected to a vulnerability at the application layer. Thus, securing these applications, or in the case of low-code, the layer that generates them, is essential for business to function smoothly.

When it comes to low-code, Wysopal notices a few key risk areas:

Proprietary libraries with hidden vulnerabilities. By their nature, low-code platforms typically use a good deal of proprietary software, making securing them a bit opaque. These systems could involve a proprietary language, proprietary libraries and proprietary frameworks. If you trust a low-code platform, you place a lot of faith in the low-code vendor to have a strong security process in place.

Some platforms do generate traditional code, which organizations can export to run on a Node.JS or Java server. Yet, this often involves a mix of code and proprietary libraries, says Wysopal. While you could run traditional static analysis on the generated code, you may miss some context.

Lack of citizen developer security training. Citizen developers have little technical knowledge, let alone DevSecOps training. If runtime applications present vulnerabilities, expecting no-code users to mitigate them is unrealistic.

API integrations. If low-code platforms expose an API or generate a web application via an API, the platform could be exposing sensitive data.

Ways to Harden Low-Code Environments

To respond to these areas, Wysopal offers some advice for securing low-code environments:

  • Perform static code analysis: Perform your own static analysis on any generated code and test for common errors. “Look at the code, and understand where it interacts with the outside,” Wysopal recommended.
  • Audit proprietary libraries: Whenever possible, Wysopal said he recommends pushing back on the vendor to question them on their application security standards and examine proprietary libraries for potential risks.
  • Verify partners: Similarly, take precautions when working with third-party partners around low-code tools. Partner tools should be carefully vetted and hold some sort of application security certification from the low-code platform.
  • Secure the API layer: Understand what APIs the application interacts with. These connections should be tested dynamically and automatically with an API scanner. Along your journey, consider these API security testing best practices.
  • Compartmentalize: Wysopal also recommends traditional isolation techniques. Consider more than just the application layer; Wysopal recommended isolating the app and running it in a virtual machine or a container.
  • Security training for citizen developers: Since citizen developers are typically not trained on appsec practices, hold security exercises to train them on fundamentals. A helpful demonstration could involve directly attacking an app to find flaws.
  • Apply the least-privilege-as-possible rule: When assigning user capabilities, whenever possible, use secure-by-default settings and allow access only to those that need it.

By applying dynamic and runtime techniques, you may catch vulnerabilities, such as a cross-site scripting error. However, if you’re working with proprietary low-code systems, you may not be able to fix it immediately! You may have to send a request to the vendor and wait for the bug to be patched. Thankfully, most technology providers take more immediate action in response to bugs than to feature requests, Wysopal noted.

Change the Attitude Toward Security

On the low-code spectrum, there is the attitude that “people aren’t really doing mission-critical applications — they’re mainly doing prototyping, or back office automations,” Wysopal said. Or, since low-code applications often do not publicly expose an application, they are deemed low-risk and fall to the bottom of the security team’s inbox and list of to-dos. “We’re still in a world where people don’t secure all applications,” he said.

However, these attitudes are a bit short-sighted, since “applications aren’t islands,” Wysopal said. Even if applications are intended for internal use only, phishing attempts could expose credentials and gain access to the network. Here, attackers could leverage sensitive resources or steal infrastructure for compute power.

Thus, all low-code users should be aware of potential threats and change habits to address these risks accordingly. However, the onus is also on low-code vendors to sufficiently arm their systems. Having a vulnerability disclosure, bounty program and an easy means to accept bug reports from white hat security researchers will be necessary to continually patch issues.

No-Code ≠ No-Bug

Low-code continues to permeate more and more digital operations, opening up novel potential for citizen developers. While the low-code movement promises impressive returns, it also brings potential risk.

To mitigate these concerns, we must level up our security understanding and evolve our approaches, Wysopal said. “Any application can have flaws and security bugs in them.” Just because you’re not writing a function in C and are relying on a visual programming model doesn’t mean you’re not introducing flaws.

Disseminating the real risk of resources and dependencies, along with helpful consumer-facing documentation, will set the superior low-code platforms apart from the rest. Otherwise, putting faith in proprietary libraries will be a tough pill to swallow.

Leave a Reply

Your email address will not be published. Required fields are marked *