Enhancing an Application Security Program: The Importance of Technology in a Maturity Model

By 
Charan Akiri
,
and
November 8, 2023

In today's rapidly evolving digital landscape, a robust software security program is no longer a luxury but a necessity. The effectiveness of security programs heavily depends on the people, processes, and technologies involved. These factors can distinguish between an application remaining secure or facing the imminent risk of a breach. 

The Purple Book Community's Scalable Software Security Maturity Model (S3M2) underscores the importance of scalability and collaborative efforts within the community. This means it's designed not just as a one-size-fits-all solution but as a versatile framework that can be tailored to suit organizations across various scales and sectors. Moreover, it actively promotes interaction with the broader software security community, fostering a culture of knowledge exchange and tapping into shared expertise to bolster security measures. 

S3M2 is structured around three core pillars of People, Process, and Technology (PPT), each having sub-categories. On the  maturity axis, the model has five levels:

  1. Ad-Hoc
  2. Proactive 
  3. Managed
  4. Optimized 
  5. Dynamic

For a more in-depth look into the dimensions and levels of S3M2 based on the PPT framework, visit this link. In this article, we'll dive into the Technology facet of S3M2.

The 5 levels of S3M2


While evaluating the maturity levels listed below, it's essential to consider that each organization may have specific elements in its program aligned with different levels. S3M2 does not mandate strict adherence to a specific maturity level, nor must every organization reach Level 5—certainly not for every aspect of software security. The optimal level for an organization will depend on various factors:

  • Support from Leadership
  • Resources, Budget
  • Regulatory Requirements
  • Timelines
  • An organization's or industry's specific requirements

The authors of S3M2 understand that context, environment, and organization objectives influence investment and implementation decisions. One organization may require a much broader toolset earlier in its development, while another may need to focus on automation as much as possible. S3M2’s groupings are generalized and stereotypical; and not meant to be mandates. Instead, consider that what has been described are the clusters of approaches and methods that a typical organization at a particular maturity level will tend to exhibit.

S3M2 views the technology domain as consisting of 2 areas:

  • Tooling
  • Secure Design 

Let's explore each one of them.

Tooling

Tooling represents the suite of tools employed across the Software Development Life Cycle (SDLC) to bolster security measures, from design and development to testing and remediation. In the nascent stages, organizations often resort to basic tools like GitLab CI or GitHub Dependabot, focusing primarily on overt vulnerabilities. As they evolve, tools become more aligned with the intricacies of the codebase and the organization's environment. By reaching an optimized maturity level, they incorporate advanced testing methods, custom rules, and policy enforcement. Ultimately, a dynamic state is achieved when tools are seamlessly orchestrated, ensuring automated end-to-end security testing and response. This progression underscores the importance of tooling evolution, bridging the gap between rudimentary practices and advanced security mechanisms aiding organizations in their journey towards fortified software security.

Ad-Hoc

Organizations employ a limited set of tools, often opting for freely available or straightforward ones like  GitLab CI™ or GitHub Dependabot®. GitLab CI or GitHub Dependabot. These tools primarily focus on vulnerabilities in inclusions, such as Software Composition Analysis (SCA). There may be sporadic use of Static Application Security Testing (SAST) with partial code coverage and ad-hoc test regimes.

Proactive

There's basic visibility into open source usage, code health, and potential external threats. This involves tools for secrets detection, SCA, SAST, and ad-hoc penetration testing. Often, only the most adept developers use these tools, and the focus is on easily understandable issues. Passive tool usage is prevalent, and there's a reluctance to "break the build." The selection and implementation of tools might happen with limited input from developers, leading to potential organizational friction.

Managed

At this level, tools are precisely tuned to the codebase, environment, and specific use cases. There's a synergy between tools and the SDLC, incorporating Dynamic Application Security Testing (DAST), Mobile Application Security Testing (MAST), Infrastructure as Code (IaC), and more. This stage sees the incorporation of automated bug tracking, risk management tools, and a dedicated testing environment mirroring production. Crucially, there's a collaborative approach between security and development teams in tool selection.

Optimized

Incorporating Interactive Application Security Testing (IAST) and standardized threat modeling workflows are hallmarks of this stage. There's a significant emphasis on policy enforcement, such as automatically rejecting non-compliant build attempts. Complementary testing methods like Fuzzing, IAST, and runtime protection mechanisms are used. Custom rules often come into play, with risk management becoming an integral component of project management.

Dynamic

The Dynamic level of AppSec maturity is achieved by automating security testing and response into an advanced end-to-end security testing process. This may involve using SOAR platform(s) to automate security tasks, integrating security testing into the development pipeline, and orchestrating security tools to work together seamlessly.

Secure Design 

Secure Design is embedding security considerations at the very core of software design, grounded in policy standards across the entire SDLC. Secure Design involves identifying, assessing, and mitigating security risks at all stages of the SDLC, from requirements gathering to deployment and maintenance. In terms of tooling, Secure Design involves using tools and techniques to automate code analysis, vulnerability scanning, configuration setup, and infrastructure assessment to identify potential risks and implement mitigation strategies.

 In its elementary form, organizations may exhibit a lackadaisical approach with limited threat modeling. As they mature, there's a progressive shift from reactionary measures to proactively embedding security into designs, making threat modeling an integral part of the process. In the managed phase, a structured process for engaging secure design expertise is adopted, ensuring security changes undergo rigorous threat modeling. Reaching an optimized state sees the standardization of threat modeling practices and an emphasis on third-party code and services that adhere to proven security standards. In its zenith, the dynamic phase, secure design becomes all-encompassing; threat modeling is integrated throughout the SDLC, and security measures are perpetually monitored and refined. This journey exemplifies the importance of a proactive approach to design, ensuring software is not only resilient against current threats but is also future-proofed against emerging vulnerabilities.

Ad-Hoc

Secure design practices are virtually non-existent. Threat modeling, if present, is informal and limited by developers' knowledge. Security requirements are typically reactionary.

Proactive

Basic threat modeling is introduced for vital designs or applications. However, there's usually a significant expertise gap in secure design.

Managed

There's a structured process for engaging in secure design expertise, with a mandatory threat modeling requirement for significant security changes. Secure design expertise becomes accessible for crucial design tasks.

Optimized

Threat modeling becomes a regularized practice with a standardized documentation form and methodology. Governance ensures model quality, and there's a structured risk rating approach. There's an emphasis on using third-party code and services with proven security controls.

Dynamic

To achieve the highest level of AppSec maturity, organizations must implement a multi-tiered secure design skill structure, integrate threat modeling into all aspects of the software development lifecycle, and automate security testing into their continuous integration and continuous delivery (CI/CD) pipelines. Security must also be monitored and responded to effectively. By implementing the right technologies and processes, organizations can achieve a Dynamic level of AppSec maturity and protect their applications and data from cyberattacks.

Summary

Application security maturity is a continuous journey rather than a fixed endpoint. One of the crucial components in this journey is leveraging the right tools, enabling us to attain desired security maturity levels quickly and efficiently.

As demonstrated by the S3M2 model, technology's role is not static; it evolves from rudimentary tools and practices to a sophisticated mesh of automation, orchestration, and stringent policy enforcement. This journey highlights the importance of continuous investment, experimentation, and evolution in the technological domain of software security. As organizations mature, the emphasis must remain on selecting the right tools, fostering collaboration, and ensuring that technology serves its purpose: safeguarding the digital assets and reputation of the organization.

Charan Akiri
Senior Application Security Engineer, Reddit