The Secure Controls Framework (SCF) release 2023.2 containscompletely new content for its Security & Privacy Capability Maturity Model(SP-CMM). This effort was conducted to help streamline and standardize maturitycriteria. If you are unfamiliar with the SP-CMM, it has been around for about 4years and is a component that is built into the SCF. The SP-CMM draws upon thehigh-level structure of the SystemsSecurity Engineering Capability Maturity Model v2.0 (SSE-CMM).
The SP-CMM’s control-level maturity criteria are designed sothat each succeeding level of maturity is built upon its predecessor.Essentially, you cannot run without first learning how to walk. Likewise, youcannot walk without first learning how to crawl. This approach to definingcybersecurity & privacy control maturity is how the SP-CMM is structured.
The SP-CMM is meant to solve the problem of objectivity in bothestablishing and evaluating cybersecurity and privacy controls. There are four(4) main objectives for the SP-CMM:
Doyou need to answer “yes” to every bullet pointed criteria under a level ofmaturity in the SP-CMM? No. We recognize that every organization is different. Therefore, thematurity criteria items associated with SCF controls are to help establish whatwould reasonably exist for each level of maturity. Fundamentally, thedecision comes down to assessor experience, professional competence and commonsense.
The following two (2) questions should be kept in mind whenevaluating the maturity of a control or Assessment Objective (AO):
While a more mature implementation of controls can equate toan increased level of security, higher maturity and higher assurance are notmutually inclusive. Froma practical perspective, maturity is simply a measure of governance activitiespertaining to a specific control or set of controls. Maturity does notequate to an in-depth analysis of the strength and depth of the control beingevaluated (e.g., rigor).
According to NIST, assurance is “grounds for confidencethat the set of intended security controls in an information system areeffective in their application.”[1]Increased rigor in control testing is what leads to increased assurance.Therefore, increased rigor and increased assurance are mutually inclusive.
The SCFConformity Assessment Program (SCF CAP) leverages (3) three levels ofrigor. The SCF CAP’slevels of rigor utilize maturity-based criteria to evaluate a control, since amaturity target can provide context for “what right looks like” at a particularorganization:
SP-CMM Level 0 (L0) - Not Performed
This level of maturity is defined as “non-existencepractices,” where the control is not being performed:
L0practices, or a lack thereof, are generally considered to be negligent.The reason for this is if a control is reasonably-expected to exist, by notperforming the control that is negligent behavior. The need for the controlcould be due to a law, regulation or contractual obligation (e.g., clientcontract or industry association requirement).
SP-CMMLevel 1 (L1) - Performed Informally
This level of maturity is defined as “ad hoc practices,”where the control is being performed, but lacks completeness & consistency:
L1practices are generally considered to be negligent. The reason for thisis if a control is reasonably-expected to exist, by only implementing ad-hocpractices in performing the control that could be considered negligentbehavior. The need for the control could be due to a law, regulation orcontractual obligation (e.g., client contract or industry associationrequirement).
SP-CMMLevel 2 (L2) - Planned & Tracked
Practicesare “requirements-driven” where the intent of control is met in somecircumstances, but not standardized across the entire organization:
L2 practices are generally considered to be “audit ready”with an acceptable level of evidence to demonstrate due diligence and due carein the execution of the control. L2 practices are generally targeted onspecific systems, networks, applications or processes that require the controlto be performed for a compliance need (e.g., PCI DSS, HIPAA, CMMC, NIST800-171, etc.). It can beargued that L2 practices focus more on compliance over security. Thereason for this is the scoping of L2 practices are narrowly-focused and are notenterprise-wide.
SP-CMMLevel 3 (L3) - Well Defined
This level of maturity is defined as “enterprise-widestandardization,” where the practices are well-defined and standardized acrossthe organization:
L3 practices are generally considered to be “audit ready”with an acceptable level of evidence to demonstrate due diligence and due carein the execution of the control. Unlike L2 practices that are narrowly focused,L3 practices are standardized across the organization. It can be argued that L3 practices focus onsecurity over compliance, where compliance is a natural byproduct of thosesecure practices. These are well-defined and properly-scoped practicesthat span the organization, regardless of the department or geographicconsiderations.
SP-CMMLevel 4 (L4) - Quantitatively Controlled
This level of maturity is defined as “metrics-drivenpractices,” where in addition to being well-defined and standardized practicesacross the organization, there are detailed metrics to enable governance oversight:
L4 practices are generally considered to be “audit ready”with an acceptable level of evidence to demonstrate due diligence and due carein the execution of the control, as well as detailed metrics enable anobjective oversight function. Metrics may be daily, weekly, monthly, quarterly,etc.
SP-CMMLevel 5 (L5) - Continuously Improving
This level of maturity is defined as “world-classpractices,” where the practices are not only well-defined and standardizedacross the organization, as well as having detailed metrics, but the process iscontinuously improving:
L5 practices are generally considered to be “audit ready”with an acceptable level of evidence to demonstrate due diligence and due carein the execution of the control and incorporates a capability to continuouslyimprove the process. Interestingly,this is where Artificial Intelligence(AI) and Machine Learning (ML) wouldexist, since AI/ML would focus on evaluating performance and making continuousadjustments to improve the process. However, AI/ML are not required tobe L5.