The Open Group Standard: Zero Trust Reference Model (Snapshot)
The Open Group Standard: Zero Trust Reference Model (Snapshot)
NOTICE
Snapshot documents are draft standards, which provide a mechanism for The Open Group to disseminate information on its current
direction and thinking to an interested audience, in advance of formal publication, with a view to soliciting feedback and comment.
A Snapshot document represents the interim results of an activity to develop a standard. Although at the time of publication The
Open Group intends to progress the activity towards publication of a Preliminary Standard or (full) Standard of The Open Group,
The Open Group is a consensus organization, and makes no commitment regarding publication. Similarly, a Snapshot document
does not represent any commitment by any member of The Open Group to make any specific products available.
This Snapshot document is intended to make public the direction and thinking about the path we are taking in the development of
the Zero Trust Reference Architecture. We invite your feedback and guidance. To provide feedback on this Snapshot document,
please send comments by email to ogspecs-snapshot-feedback@opengroup.org no later than April 30, 2024.
This Snapshot document is valid through April 30, 2024 only.
For information on joining the Security Forum, please send email to https://www.opengroup.org/join-forum.
Copyright © 2023, The Open Group
The Open Group hereby authorizes you to use this document for any purpose, PROVIDED THAT any copy of this document, or any
part thereof, which you make shall retain all copyright and other proprietary notices contained herein.
This document may contain other proprietary notices and copyright information.
Nothing contained herein shall be construed as conferring by implication, estoppel, or otherwise any license or right under any patent
or trademark of The Open Group or any third party. Except as expressly provided above, nothing contained herein shall be construed
as conferring any license or right under any copyright of The Open Group.
Note that any product, process, or technology in this document may be the subject of other intellectual property rights reserved by The
Open Group, and may not be licensed hereunder.
This document is provided “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED,
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE, OR NON-INFRINGEMENT. Some jurisdictions do not allow the exclusion of implied warranties, so the
above exclusion may not apply to you.
Any publication of The Open Group may include technical inaccuracies or typographical errors. Changes may be periodically made to
these publications; these changes will be incorporated in new editions of these publications. The Open Group may make
improvements and/or changes in the products and/or the programs described in these publications at any time without notice.
Should any viewer of this document respond with information including feedback data, such as questions, comments, suggestions, or
the like regarding the content of this document, such information shall be deemed to be non-confidential and The Open Group shall
have no obligation of any kind with respect to such information and shall be free to reproduce, use, disclose, and distribute the
information to others without limitation. Further, The Open Group shall be free to use any ideas, concepts, know-how, or techniques
contained in such information for any purpose whatsoever including but not limited to developing, manufacturing, and marketing
products incorporating such information.
If you did not obtain this copy through The Open Group, it may not be the latest version. For your convenience, the latest version of
this publication may be downloaded at www.opengroup.org/library.
2 Definitions................................................................................................................. 5
2.1 Zero Trust ....................................................................................................... 5
2.2 Zero Trust Architecture .................................................................................. 5
The Open Group is a global consortium that enables the achievement of business objectives
through technology standards. Our diverse membership of more than 900 organizations includes
customers, systems and solutions suppliers, tool vendors, integrators, academics, and consultants
across multiple industries.
The mission of The Open Group is to drive the creation of Boundaryless Information Flow™
achieved by:
• Working with customers to capture, understand, and address current and emerging
requirements, establish policies, and share best practices
• Working with suppliers, consortia, and standards bodies to develop consensus and
facilitate interoperability, to evolve and integrate specifications and open source
technologies
• Offering a comprehensive set of services to enhance the operational efficiency of
consortia
• Developing and operating the industry’s premier certification service and encouraging
procurement of certified products
The Open Group publishes a wide range of technical documentation, most of which is focused on
development of Standards and Guides, but which also includes white papers, technical studies,
certification and testing documentation, and business titles. Full details are available at
www.opengroup.org/library.
This Document
This is a Snapshot document of what is intended to become the Zero Trust Reference Model
Standard. It is being developed by The Open Group.
This document builds on The Open Group Snapshot: Zero Trust Commandments to provide the
basic concepts for and architectural building blocks of a Zero Trust Reference Model. The
document also describes the relationships among these architectural building blocks.
COBIT is a registered trademark of the Information Systems Audit and Control Association
(ISACA).
Microsoft is a registered trademark of Microsoft Corporation in the United States and/or other
countries.
All other brands, company, and product names are used for identification purposes only and may
be trademarks that are the sole property of their respective owners.
The Open Group gratefully acknowledges the contribution of the following people in the
development of this document:
• Didier Beyens, DXC
• Chris Carlson, C T Carlson, LLC
• Tony Carrato, The Open Group Invited Expert
• Simon Cross, (formerly of) BizzDesign
• Mats Gejnevall, Biner Consulting
• Jim Hietala, The Open Group VP, Security & Business Development
• Nikhil Kumar, Applied Technology Solutions, Inc.
• Mike Leuzinger, Nationwide
• John Linford, The Open Group Security Forum & OTTF Director
• Carmichael Patton, Microsoft
• Andy Ruth, Sustainable Evolution, Inc.
• Mark Simos, Microsoft
• Andras Szakal, The Open Group VP & CTO
• Altaz Valani, (formerly of) Security Compass
• Steve Whitlock, The Open Group Invited Expert
• Hasan Yasar, Carnegie Mellon SEI
(Please note that the links below are good at the time of writing but cannot be guaranteed for the
future.)
[C119] The Open Group Standard for SOA Reference Architecture (C119), published
by The Open Group, December 2011; refer to: www.opengroup.org/library/c119
[C20A] The Open Group Standard for Risk Analysis (O-RA) (C20A), published by The
Open Group, November 2021; refer to: http://www.opengroup.org/library/c20a
[C20B] The Open Group Standard for Risk Taxonomy (O-RT) (C20B), published by
The Open Group, November 2021; refer to:
http://www.opengroup.org/library/c20b
[C220] The TOGAF® Standard, 10th Edition (C220), published by The Open Group,
April 2022; refer to: www.opengroup.org/library/c220
[Cheswick, 1990]
The Design of a Secure Internet Gateway, by Cheswick, B, published April
1990; refer to: https://cheswick.com/ches/papers/gateway.pdf
[G202] SOA for Business Technology, The Open Group Guide (G202), published by
The Open Group, February 2020; refer to: www.opengroup.org/library/g202
[ISO 73:2009] ISO Guide 73:2009: Risk Management – Vocabulary, published by ISO,
November 2009; refer to: https://www.iso.org/standard/44651.html
[NIST, 2012] NIST SP 800-61 Rev. 2: Computer Security Incident Handling Guide, published
by NIST, August 2012; refer to: https://csrc.nist.gov/publications/detail/sp/800-
61/rev-2/final
[S230] Zero Trust Commandments (Snapshot) (S230), published by The Open Group,
August 2023; refer to: www.opengroup.org/library/s230
[S231] The Open Group Standard: Security Principles for Architecture Snapshot V1.0
(S231), published by The Open Group, October 2023; refer to:
www.opengroup.org/library/s231
[W124] Jericho Forum® Commandments (W124), published by The Open Group, May
2007; refer to: www.opengroup.org/library/w124
1.1 Objective
The objective of this document is to provide a set of models to support the core concepts of Zero
Trust. This includes the ability to develop and implement a strategy and the ability to incorporate
and define a core security framework (including Risk and Information Security Management). The
document also defines a capability-based Technology Reference Model (TRM) to help define and
implement Zero Trust Architectures (ZTAs), as well as the supporting philosophy.
The subject of this Snapshot document is the specification of a Zero Trust Reference Model, an
aggregate of a set of models associated with ZTA.
This Snapshot document is intended to make public the direction and thinking about the path we
are taking in the development of the Zero Trust Reference Model. We invite your feedback and
guidance. To provide feedback on this Snapshot document, please send comments by email to
ogspecs-snapshot-feedback@opengroup.org no later than April 30, 2024.
1.2 Overview
Zero Trust is a holistic security capability for the information security of a modern Digital
Enterprise. It includes a strategy, an approach for how security in the modern world should be
done, as well as a framework on how and what to do.
This Snapshot document presents a proposal for a standard that defines and describes all the
models required for the development and implementation of a Zero Trust Strategy, including the
Zero Trust Implementation Model, the Zero Trust Risk Model, The Zero Trust Information
Security Management Model, and the Zero Trust Reference Model, with their underlying
architecture philosophy and principles.
The Reference Model (called the “Zero Trust Reference Model”) provides a single point of
reference for multiple stakeholders from across the organization to align on what ZTA means, and
how it comes together. It supports and implements the core tenets of Zero Trust: data-centricity,
asset-centricity, adaptive access control (policy-driven access control), modern identity
management across all identities (including digital identity), and security zones. It also enables
organizations to execute the strategy, integrate with business processes, and establish a model for
the sustained use of the core concepts.
1.3 Conformance
This is a Snapshot document, not an approved standard. Do not specify or claim conformance to
it.
1.5 Terminology
For the purposes of the Zero Trust Reference Model (Snapshot) the following terminology
definitions apply:
May Describes a feature or behavior that is optional. To avoid ambiguity, the opposite of
“may” is expressed as “need not”, instead of “may not”.
1A formal harmonization initiative within The Open Group Security Forum will clearly define and describe how to use the Security
Principles for Architecture, the Security/Risk Reference Architecture, the Zero Trust Reference Model, and the Zero Trust
Commandments Snapshots, ensuring consistency and avoiding duplication and contradictions in future iterations of all documents.
For the purposes of this document, the following terms and definitions apply. The terms in The
Open Group Portfolio of Digital Open Standards – Glossary and Roles (Snapshot) [S222] and
Merriam-Webster’s Collegiate Dictionary should be referenced for terms not defined in this
section.
[Source: S230]
[Source: S230]
Zero Trust provides a modern approach to Information Technology (IT) security, made necessary
by changes in how business is done and ever-changing security risks. With growing international
adoption of Zero Trust and the rapid migration of organizations to digitization, enterprises must
understand Zero Trust and provide a structure for it.
Zero Trust updates traditional IT security architecture and operations to better support
contemporary business models. Traditional IT environments operated primarily inside of a private
network and infrastructure boundary. This model assumed that the private network confers trust
on assets hosted on it, but this assumption has proven false. Contemporary business environments
are composed of an ecosystem of digital products being accessed by many stakeholders across any
physical and internet locations. Zero Trust replaces the traditional IT security approach and
provides security and risk management without relying on the assumption of a trusted network.
Many aspects of Zero Trust are not new concepts. Concerns with a network perimeter-centric
approach can be traced as far back as 1990 (which were famously captured by Bill Cheswick’s
quote of “…a sort of crunchy shell around a soft, chewy center” [Cheswick, 1990]). The Jericho
Forum formalized this shift away from the network perimeter-centric model starting in 2004,
which was documented in the Jericho Forum Commandments [W124]. This Forum has since
merged into The Open Group, making this work a foundational pillar for Zero Trust.
Today, Zero Trust is recognized as an imperative by many organizations from the US White
House3, NIST4, and the US Department of Defense (DoD)5 to commercial organizations such as
Microsoft, Google, and countless security companies. This reference model has been informed
by this work and is intended as generalized guidance that applies to an organization’s entire
technical estate across IT, OT, and IoT types of assets.
This model is designed to mitigate both known and emerging security risks.
The definition for Zero Trust from The Open Group Snapshot: Zero Trust Commandments [S230]
can be expanded and divided into three perspectives.
Zero Trust:
1. Is an information security approach that focuses on the entire technical estate – including
data/information, APIs, and Operational Technology/Industrial Control Systems –
throughout their lifecycle and on any platform or network.
2. Provides a security framework to protect assets anywhere based on asset-centric and data-
centric security, policy-driven access controls, modern incident detection and response,
modern identity management, and security zones/domains.
3. Enables organizational flexibility, agility, and adaptability while continuing to provide the
same (and often stronger) security assurances of confidentiality, integrity, and availability
for business assets.
Zero Trust is implemented through a comprehensive strategy and provides a security framework
based on asset or data-centric security, policy-driven controls, modern identity management, and
security zones/domains. It also includes the operational aspects that allow organizations to adopt
and apply Zero Trust in a holistic manner. This definition, along with the operational aspects form
foundational concepts for Zero Trust and should be considered when thinking about Zero Trust.
Zero Trust is the Information Security Architecture of the digital era. Organizations operating in
the modern world, especially but not limited to hybrid cloud environments, should consider Zero
Trust as the de facto approach for their information security. This involves executing a Zero Trust
strategy, typically aligned with the business and IT strategy (if they exist) and establishing a
runtime capability, leveraging Zero Trust capabilities.
In keeping with modern drivers of agility, flexibility, and velocity, Zero Trust enables modern
organizations to operate and execute their business with confidence.
An architectural “point of reference” around which organizations can align. This TRM is described
in terms of capabilities and Architectural Building Blocks (ABBs) and the standards around them.
Where such standards do not already exist, new standards are identified and described at a high
level in this document or are defined in detail in ancillary documents. The TRM also provides the
baseline around which Zero Trust Architectures (ZTAs) are defined and realized.
Models which help organizations define and execute a pathway to implementing Zero Trust,
including:
• Devising and implementing a strategy (strategic implementation models)
• Risk assessments leveraging quantitative risk analysis (e.g., the Open FAIR Standards)
• Security Management
The models in this document cover all the major aspects of Zero Trust, helping guide an
organization’s Zero Trust mission, vision, strategic roadmap, architecture, implementation,
operation, and governance of Zero Trust. This is illustrated in Figure 1.
This document is composed of four specific models that are described in this document:
• The Zero Trust Implementation Model (Section 5.1) (also referred to as the 3-Pillar
Model6) lays out the structure around which to develop and implement Zero Trust
strategies
— This defines how to develop, implement, govern, and run a Zero Trust Strategy and
capability
Use the Zero Trust Implementation Model to provide a framework around the
development of Zero Trust strategies and their implementation. Apply the tailoring as
described above to develop governance frameworks.
• The Zero Trust Information Security Management (ISM) Models (Section 5.2) lay the
foundation for the operational use of Zero Trust, addressing ISM issues
6Note that this model is adapted from the Service Oriented Architecture (SOA) for Business Technology Guide [G202] and adds Zero
Trust aspects as relevant. It provides a framework and reuses well understood concepts.
Use the reference model to create solution architectures by tailoring the Zero Trust Technology
Reference Model to your technology estate. Use the detailed description of the capabilities and the
ABBs in Chapter 5 to map to specific standards within this document that might apply to your
solution architectures. The Zero Trust Implementation Model provides a model to develop and
implement a Zero Trust Strategy, leveraging the Zero Trust Reference Model to create Target and
Transition State Solution Architectures.
This model should be tailored to the unique requirements of an organization including business
drivers, line of business (industry/domain), organizational business and operating models, local
regulatory controls, and the organization’s existing technical estate and strategy. This tailoring
should also consider intangibles, such as organizational maturity level, institutional knowledge,
and culture. The organization should then determine capabilities, technologies, and products and
their prioritization, to form Solution Reference Models, which can inform architectural decisions
resulting in technology, product, and tool selection, and solution implementation. Organizations
can also use these models to establish a common taxonomy and terminology across the various
entities within the organization. Note that the Zero Trust Reference Models metamodel (Section
5.1.1) defines a metamodel for Architects (both security and technology) to use, including the
concepts underlying ABBs, Solution Building Blocks (SBBs), capabilities, etc., and the relations
between them.
Other standards bodies can also use this to develop complementary standards, establish
interoperability, and support a consistent industry.
Senior security leaders such as Chief Information Security Officers (CISOs), Chief Information
Security Architects (CISAs) and Chief Information Officers (CIOs) can use this document to help
set organizational vision, develop, and implement roadmaps, and set up organizational units and
governance frameworks.
Security Directors and managers, IT Directors and managers, and other leaders within security
and technology teams can use this document to drive direction and implementation of Zero Trust
Roadmaps, align teams, and assess the impact of Zero Trust.
Any other leaders that are sponsoring, planning, and supporting security modernization initiatives
and other Zero Trust-related activities can use this document to determine impact, establish a
common understanding, and work on the development and implementation of a Zero Trust
Roadmap.
Security architects operating at the enterprise level can use this document to support the business
structure of the organization, including the ability to operate quickly and flexibly, to guide
procurement of new tools (largely for security), which ensures that those tools support the
organizations Zero Trust intent, and to support that Zero Trust intent in organizational governance
and risk management activities.
Security architects can use this document to help develop and implement Zero Trust Roadmaps.
Security architects can use this document when working with infrastructure, application, network,
and other technical teams, as well as teams which enable security as part of business strategy and
operation.
Application Architects and Delivery Teams, including Technology and IT teams involved in
supporting activities such as Development, Security, and Operations (DevSecOps) can use this
document to help establish a common shared understanding of the details of Zero Trust and the
development of cross-cutting and reusable assets.
Information Architecture teams involved in data management, governance, and privacy can use
this document along with Security and Enterprise/Solution Architects, Business leaders, and
product owners to help identify and define data assets and their lifecycle and business value.
Security and IT engineers/developers can use this document to help establish a common shared
understanding of the details of Zero Trust and the development of cross-cutting and reusable
assets.
Security and IT analysts can use this document to help establish a common shared understanding
of the details of Zero Trust, the development of cross-cutting, reusable assets, oversight, and
governance to ensure that security is covered in general and in the context of the various
viewpoints.
Audit and compliance teams can use this document to ensure on-demand audit and compliance
with different risk and regulatory policies and controls.
Risk teams can use this document to define and ensure compliance with controls, and to define
the business value of assets, and the evaluation and assessment of risk and appropriate protection
of assets.
Testing teams can use this document to determine testing regimes, tools, and capabilities that need
to be supported and provide input into the Zero Trust roadmap and its implementation.
Organizations developing products for the digital enterprise, both for internal use and sales to other
organizations can use this document to develop conformant products that support the capabilities
and ABBs.
Other standards bodies and institutions can use this document to develop complementary standards
and to leverage a shared industry understanding of the terminology and vocabulary of Zero Trust.
Note that the different classes of users of this document may use the document in more ways than
described above.
Zero Trust allows organizations to operate securely on any network at any time in a state of
assumed compromise (assumed breach).
A Zero Trust capability includes the architecture that provides the foundation for achieving this
vision, a strategic implementation approach, an operational capability, and an ability to assess risk
and opportunity.
The implementation of security (including the simple and direct approach of Zero Trust) is
complicated by the realities that security risks affect the entire organization (they are not contained
to any given system or business unit) and that the measures to mitigate security risk must be part
of everyone’s job.
While Zero Trust is similar in mission to previous approaches to information security, it has a
different fundamental assumption of trust. Instead of relying on an invalid assumption that an
organization’s network is (or could be) safe and trustworthy, Zero Trust assumes that assets are
connected to and communicating over open (and potentially hostile) networks, like the Internet,
and requires securing them accordingly. This shift is depicted visually in Figure 2.
This change in assumption leads to asset and data-centric security approaches that protect each
asset individually, require explicit validation of all claims of trust (rather than relying on implicit
trust of network location), and limit the damage (or blast radius) from any asset that is
compromised.
This also shifts security from a technology-centric approach (do you have XYZ technology?) to a
capability – and outcome-centric approach (are ABC assets protected against current top threats?).
An enterprise security model based on Zero Trust assumes that an enterprise network has already
been penetrated, or could be at any time. This requires security architectures and controls to be
designed and prioritized differently. While Zero Trust embraces existing investments into network
security, it goes well beyond this single control type.
Zero Trust:
• Moves to a policy-driven model that focuses security controls on individual assets rather
than on network egress points
This tailoring of security to each asset type naturally reduces the threat surface (or attack
surface) and the blast radius of damage for those assets if they are compromised
• Prioritizes using the intrinsic business value of assets to ensure the strongest protections
are applied to the most valuable assets, ensuring security resources are utilized
commensurate to asset value
• Enables continuous monitoring and subsequent improvements to security controls and
assurances using all available telemetry, intelligence, and data
These asset-centric Zero Trust capabilities increase the agility of the organization by ensuring the
assets themselves are always protected in any situation or configurations. These protections remain
even as business processes, use cases, and configurations change to meet the evolving needs of
the digital ecosystem. The data-driven approach of Zero Trust also allows rapid audit, compliance,
and risk capabilities, further supporting business agility to rapidly enter new markets and reduce
organizational risk.
The Digital Enterprise and the environment in which it operates are ever evolving, ever more
complex, and constrained by the need to deliver faster and within resource constraints. Thus, Zero
The rate of changes in the threat environment, technical platforms, and market preferences is
increasing, driving the need for Digital Business Transformation, cloud technology
transformation, and a Zero Trust security transformation to keep up with them.
Figure 3 provides a view of a Zero Trust Architecture that captures the key essence of the Zero
Trust Transformation.
Zero Trust is transformative and requires cultural change, technology change, process changes,
technical operating model changes, and skills changes. In its operation, Zero Trust is also more
dynamic than previous security models and approaches, requiring the adoption of Agile roadmaps
that can adapt to the rapid technology, business, and cloud changes. Zero Trust is not executed as
a single big overnight change, but as incremental changes that steadily realign many aspects of the
strategy and organization to the new direction.
Zero Trust breaks down silos separating security and technology teams and activities. Zero Trust
drives alignment across different disciplines and organizational units such as business, security,
technology, risk, and compliance. The Open Group Snapshot: Zero Trust Commandments [S230]
provides a foundation for guiding that change. Organizations shall use the Zero Trust
Commandments to define their Zero Trust journey and provide a shared vision of Zero Trust to all
stakeholders in the organization, including business leaders.
Zero Trust Governance introduces security architecture and threat intelligence as governance
functions to drive informed decisions across systems.
This Zero Trust function is a greatly expanded and integrated version of vulnerability scanning.
Posture management enables a proactive holistic security approach that allows the organization to
burn down the ‘technical debt’ of weak security practices that have accumulated over 30+ years
Posture management includes both an inside-out and outside-in view covering internal scanning
and external attack surface management. Internal scanning supports an inside-out approach uses
management tools to monitor the security status of the organizations posture such as a Cloud
Security Posture Management (CSPM) tool that continuously reports on and makes
recommendations to improve security posture. External Attack Surface Management (Outside-in
scanning) – used tools that monitor the security posture and attack surface of the organization from
the internet (mimicking an attacker’s view of the organization). Tools like External Attack Surface
Management (EASM) products monitor what the digital footprint of the organization looks like
across their platforms, websites, brands, multiple cloud types and providers, on-premises
datacentres, mobile, social, third parties, and more.
4.1.4 IT Operations
IT operations are responsible for managing enterprise-wide technical resources in the technical
estate. Any changes to the production environment from posture management or security
operations are done in partnership with IT Operations. Zero Trust introduces a close integration
with Security Operations for rapid incident response and recovery (mitigating realized risk) and
with posture management for prevention (mitigating potential risk).
Zero Trust Data Governance requires clear data ownership, stewardship, and accountability.
Adaptive access control is an updated approach to enable and secure access to any type of asset
(resource) regardless of geographic or network location. Adaptive access control provides
consistent access policy enforcement across assets and locations that adapts to dynamically
changing factors such as threat intelligence, user behavior patterns, and more. This approach also
enables rapid Agile updates to access control policies based on changes to business requirements,
the technical estate, and the threat environment.
This enables building and enforcing access policy that is informed by the organization’s risk
appetite and continuously changing threats in real-time.
Zero Trust transforms multiple aspects of the complex function of security, so this section
introduces the models that give an overview of the Zero Trust capabilities and components, the
recommended strategic implementation approach, how to integrate it into other organizational
functions and capabilities, and then the operational aspects of Zero Trust.
Figure 4: Zero Trust Models in the Zero Trust Reference Model Standard
This section presents the 3-Pillar Model and its key components. Details about how to apply it
will be covered in the eventual Zero Trust Practitioners Guide. Practitioners can use this Zero
Trust Implementation Model to establish and execute a Zero Trust Roadmap even in the absence
of the Zero Trust Practitioners Guide.
Figure 5 shows the three pillars of the model: Strategy, Operational, and Operating Model.
7The term “risk” in this document utilizes the definition from the Open FAIR Body of Knowledge: “Risk is the probable frequency
and probable magnitude of future loss.” Where discussion of potential gain occurs, the term “opportunity” is instead used for
consistency and to delineate the concepts.
The Strategy Pillar defines the development of the roadmap by identifying the mission, vision,
and goals and mapping them into Business Capabilities. Capabilities in the context of this model
can be divided into business, technical, security, and Zero Trust capabilities. Capabilities may be
supported by people (organization), processes, and technology.
The primary outcome of the Strategy Pillar is a tailored, capability-centric roadmap. Roadmaps
are typically broken down into phases, with metrics and maturity levels helping tailor and tune the
roadmap over time. They are usually highly accurate in the short-term, have medium level of
fidelity and confidence in the mid-term, and represent high level directional guidance for the long-
term. Roadmaps link the delivery of these capabilities together and are updated regularly,
providing alignment with an evolving business, technical, regulatory, and security landscapes.
Note that as the organization develops its roadmap, legacy applications will typically use
compensating controls while they migrate to a Zero Trust Architecture (ZTA).
The Operational Pillar defines the execution of the roadmap and the tailoring of it to organizational
or departmental constraints, helping create solution architectures, map the people and
organizational units, the business processes, and build out the roadmap while the organization is
running the business.
The Operating Model Pillar provides governance and change management, helping establish key
elements of a digital culture – a culture of continuous learning, change, and improvement. It runs
simultaneously with the other two pillars, forming communications, decision rights, guardrails,
continuous learning, and other elements of a Zero Trust enterprise. The organizational operating
model forms a filter to assess the organization from a culture and business operating model
perspective and determine spending, business model, and cultural drivers. This in turn determines
the approach to take in the implementation of the roadmap.
Note: People-centricity is an important element of Zero Trust. Because mitigating security risk
is part of everyone’s job and security risks affect the entire organizations, Zero Trust
requires that organizations adopt organization-wide cultural elements and processes to
8
The next version of this document will provide more details on this content.
Scorecard level metrics are KPIs intended for senior leaders and boards, represented as a
scorecard, typically framed as a variant of a Kaplan Scorecard Model [Kaplan & Norton].
Scorecard changes related to Zero Trust must align to both the Zero Trust strategy and the
enterprise business strategy.
9
The next version of this Snapshot will provide more details on this content.
• Zero Trust Program Status and Progress – creation of new Zero Trust metrics aligned to
the Zero Trust mission, vision, and goals (described below)
Illustratively, these could provide measurements of how well the organization is able to
prevent risk or to respond to and remediate risk, and how well security enables business
processes and productivity goals.
While the details of the “how” will be covered in the eventual Zero Trust Practitioners Guide, a
short summary is:
The mission, vision, and goals are established early to ensure the organization aligns around an
aspirational mission and a vision to get there.
Developing a Zero Trust Mission, Vision, and Goals aligned to the organization’s mission helps
ensure that Zero Trust and Security teams are aligned to the context of the organization and its
business drivers. These guide the Zero Trust journey while incorporating Zero Trust standards and
capabilities and must be kept in mind when launching a Zero Trust.
The Zero Trust vision is usually time-limited and has specific constraints. In the case of Acme
Healthcare, it could be: “Our vision at Acme is to be a leader in total perfect health by improving
the lives of our members and the quality of health of their communities”. The associated Zero
Trust vision might be “being a leader in delivering the services in a secure manner in any
environment, with the highest level of availability, on any channel, with the ability to support the
Agile addition of deletion of partners and capabilities into the organizational ecosystem, and in an
environment of assumed breach, in alignment with Acme Healthcare’s Digital Transformation and
Cloud Migration initiatives”. Enabling the business vision but focused on Zero Trust. Ensuring
alignment with both the organizational and technology strategies.
The vision is then decomposed into more specific Zero Trust goals associated with specific metrics
(usually KPIs). The associated goals could then be “the ability to support Health Insurance
Portability and Accountability Act (HIPAA) compliance for the delivery of data on any channel,
and on any platform, in an environment of assumed breach”, “the ability to localize and limit the
impact of a breach to 99% of Acme’s component systems, with complete isolation of the rest”,
“the ability to recover from a disaster or breach in xxx time”, “the ability to securely share data
with our partners in an Agile manner in 2 years”, and “the ability to support the addition or removal
of 90% of Acme’s partners in a secure manner with full verifiability in a hybrid cloud environment
in 3 years”.
These illustrative mission, vision, and goals must be adapted and tailored to the individual
organization’s business and technology strategies. Use the Operating Model’s “business operating
models” to help focus on prioritization and approach keeping how an organization focuses
business structure and philosophy. This is critical for success as it focuses on both the spend,
organizational structure, and leadership support.
The capabilities enable achieving the defined goals. There may be different ways to achieve these
goals, and these will usually be curated by the organization’s business and technical environment,
its operating model, ongoing strategies, etc. Roadmaps will achieve goals using these capabilities,
typically in the form of multiple strategic initiatives. The term used to describe these initiatives
may vary based on the organizational process model (e.g., in an Agile organization, they may be
called programs or epics). These initiatives may then be associated with both metrics and maturity
levels to ensure they are able to track performance improvements.10 They will develop metrics
such as KPIs or OKRs to assist with this. Typically, OKRs provide a clear actionable model, while
KPIs are used to assess high-level organizational goals.
Illustratively, a Zero Trust capability may be data-centricity and data protection. In the Acme
context, in order to meet the goals of “the ability to securely share data with our partners in an
Agile fashion in 2 years” and “the ability to support the addition or removal of 90% of Acme’s
partners in a secure manner with full verifiability in a hybrid cloud environment in three years”,
the Zero Trust approaches of data-centricity, asset-centricity, security zones, and blast radius
reduction might be followed. Usually, a combination of these is used. For example, a roadmap
might have initiatives such as: “establish policy based adaptive access control and identity for 90%
of participants in the technical estate (ecosystem) in the Acme enterprise in one year” and
10 This Snapshot does not address a maturity model or assurance framework for Zero Trust.
The maturity measure may be the “adoption of one or more data-centric protection mechanisms
by the organization”, while metrics may be that 80% of the organization has either eliminated,
obfuscated, or established format preserving encryption of its data, and the rest is encrypted at
rest. For an illustrative OKR in this context, the objective would be obfuscation or format-
preserving encryption across the enterprise, and a key result would be 80% obfuscation or format-
preserving encryption. For a Zero Trust SOC, the organizational objective might achieving the
lowest Mean Time To Acknowledge (MTTA) (how long it takes to start work) and Mean Time
To Remediate (MTTR) (how long it takes to remove attacker access) in a particular geography or
industry, and the key result would be that the MTTA incidents should be less than “X” number of
minutes (say 10 minutes) and the MTTR incidents should be within “X” hours (say 12 hours).
Never use security operations measurements punitively. Security operations metrics shall never
be punitive, or they are likely to backfire (or never make their way into the roadmap). All security
operations metrics can be impacted by external factors not under control of the organization such
as recent vulnerabilities, attacker competency and skill, and what attack tools and automation are
available to the attacker. If people are held accountable for forces they cannot control, they have
gained an incentive to skip the work and lie to their leadership. Metrics shall be used to provide
feedback and be aspirational as OKRs are. They shall be used to train and educate people to
improve operational performance.
Zero Trust initiatives involve both significant investment and have a broad, usually enterprise-
level impact on an organization. It is important to be able to incorporate the organization’s business
Zero Trust operating model into determining what to focus on when developing the roadmap. The
operating model is covered in more detail in the review of the operating model pillar.
A metrics framework and maturity measures are used to determine the extent of success and fine-
tune requirements towards evolving organizational needs. Metrics frameworks for Zero Trust will
be elaborated on in the Zero Trust Practitioners Guide and are out of scope for this document.
Maturity models may separately be used to help define high level measures of success. OKRs are
used in Agile delivery organizations to tune organizational delivery at the unit level.
The roadmap is based on business, technical, security, regulatory, and Zero Trust capabilities. It
provides quantified targets, initiatives with the assignment of resources, and alignment with
business, technical, security, and regulatory disciplines, establishing the people skills (through
developing a culture of continuous learning) and governance framework. It leverages these
elements from the operating model pillar, using the operating model viewpoint to focus the
roadmap. The operating model helps determine what leadership focuses on and considers to be
important. This helps identify priorities, culture, and implications.
The roadmap can (and should) be viewed with a relatively high level of confidence in the short-
term, medium level of confidence in the mid-term, and directional guidance (low level of
confidence in specific details) for the long-term.
Because Zero Trust introduces changes to existing operations, think about it like changing the
engine or seats while flying the plane. Organizations have to run and grow the business while
executing the Zero Trust Roadmap, typically while simultaneously integrating (and harmonizing)
with and executing a digital business transformation and a cloud technology transformation.
The Operational Pillar takes the Zero Trust Reference Model and applies organizational, business,
regulatory, and security environments, constraints, and limitations to translate them into an
operational Zero Trust solution architecture. It applies the Zero Trust Roadmap to the
organization’s technical estate, culture, regulatory and risk context, and Process and Governance
Framework, and translates those initiatives and other elements into actionable updates to security
components that are applied to the various business and technical elements in the organization.
This allows the organization to execute the roadmap while establishing the governance framework
and solution architecture, to create an organizational evolution towards Zero Trust maturity.
The Zero Trust Operational Pillar is composed of business capabilities, processes, and functions,
as well as the technical estate that implements it. Business functions can be a combination of the
people, organizational structures, business capabilities, business processes, and the business
services provided.
What makes the Operational Pillar for Zero Trust different is that attributes that drive the
implementation of a Zero Trust Roadmap conform with the Zero Trust Commandments and
Reference Model. This creates consistency with both security and risk imperatives as well as the
impacted business elements – processes, functions, regulatory environment, business ecosystem,
business environment, and the existing technical estate. Zero Trust initiatives also tend to influence
how business is done, particularly the technology, risk, and security domains.
The resulting Zero Trust Solution Architecture and the updates to business capabilities, functions
and processes, and the technical estate are typically cross-cutting and not bound to individual
components or functions.
Additional details on executing the Operational Pillar will be covered in the eventual Zero Trust
Practitioners Guide.
For it to succeed, the Zero Trust Operating Model must be adapted to the organizational style of
the organization – a governance model designed for a unified organization with a single Chief
Executive Officer (CEO) and centralized IT processes will not work for a diversified organization
with multiple independent business units with their own CEOs and local technical and security
functions.
Trying to implement or develop a roadmap that does not take these factors into consideration
usually leads to failure. The Zero Trust Operating model uses the 4-Quadrant Model depicted in
to focus and tailor all the other components of the pillar – the governance framework, the guiding
principles, commandments, etc.
The Operating Model is based on the structure of the business and how it is organized and
prioritized. The four quadrants are:
1. Coordination: unique businesses with a need to know each other’s transactions.
2. Diversification: independent business units with different customers and expertise.
3. Unification: single businesses with global data.
4. Replication: independent, but similar, often competing business units.
The Zero Trust Practitioners Guide will tie together the various models to help organizations
develop and implement their Zero Trust journey.
Zero Trust does not require or define a specific information security management approach and is
compatible with standard definitions, including Open Information Security Management Maturity
Model (O-ISM3) [C17B].
The organization’s ISMS should be tailored to the mission, needs and objectives, security
requirements, processes, and the size and structure of the organization. Each company’s ISMS is
expected to change over time.
Any “New Incident” is managed by the Security Operations team, who performs an “Incident
Response”.
For major or new/novel incidents, a joint team effort should perform a “Root Cause Analysis” to
identify how to best mitigate risk (by identifying quick fixes for immediate effect or other
mitigations that require more analysis).
A “New Insight/Learning” can take multiple forms including a new vulnerability (such as
Log4j™), a new external incident (such as SolariGate), a new loss scenario (such as extortion or
ransomware), or a new industry best practice. These may require mitigating the risk, proactively
looking for a compromise that was previously missed (“Threat Hunting”), or both. Any threats
that are found with threat hunting (e.g., attackers that previously evaded standard detections) will
follow Incident Response processes (defined as Rapid Incident Response under Asset-Centric
Security Operations). Note that the definition of the TRM is addressed in Chapter 5.
These actions will improve security by continuously integrating learnings that help the
organization adapt to the external environment.
Information security management establishes a structure based on risk and controls and the manner
of executing them. In short, it uses risk and core drivers – security, business, risk, technology – to
determine the core governance framework required to operate the organization, especially where
these areas work together to provide the capability.
The primary function of the policy management system is translating risk in the risk register into
clear standards and requirements and into actionable control procedures that technical teams can
implement to mitigate the risks.
The work of managing policy primarily comes from the ongoing maintenance and implementation
of information security documentation, particularly the control standards. The first step in
managing change to security controls designed to mitigate a particular risk is to model the threat
surface and determine the associated controls, policies, and procedures.
Illustratively, the business addition of a new capability for vendor integration might result in the
need to support streaming as a channel, with multiple endpoints. Multiple protocols might be
involved, as well as integrations with different platforms. This now results in a threat surface area
that might vary based on the stream, its integration point and technology, etc.
Managers responsible for software development, including purchased software, need to:
• Integrate their systems with security infrastructure services
• Ensure the robustness of their application by applying a secure development process
All organizations must make decisions about relevant organizational risk. This requires
documenting and consistently executing the processes behind these decisions. Figure 13 depicts
an example of various governance functions within an organizational that are relevant in helping
an organization identify and prioritize organizational risks and business investment to mitigate
them.
Organization type and size will impact the formality of having established risk and policies teams
or councils – Small and Medium-Sized Businesses (SMBs) may have the same individual
representing multiple business functions, while larger organizations may have a formally defined
group tasked with individual responsibilities.
The Threat Response Management subprocess depicted in Figure 14 is from NIST SP 800-61 Rev.
2: Computer Security Incident Handling Guide [NIST 2012] that is the basis of incident
management at many organizations.
Threat analysis, response planning, and community communications are performed by the Manage
Incidents function. Recovery Management capabilities are performed by the Manage IT
Infrastructure Security function.
These details illustrate the difference between Manage Incidents, which is in a constant ready and
reacting state, and Manage IT Infrastructure Security, which operates to standard IT operations
processes.
The design/build and run operate aspects of Posture Management are covered in Section 6.2.6 and
Section 6.4.6. From an information security management perspective, IT infrastructures processes
must be applied to operate and manage change for security components. Most change follows
planned change cycles, incorporating requirements from security initiatives. Some changes arise
as urgent changes funded and approved by Manage Risk as required to respond to infrastructure
vulnerabilities identified by Manage Assessments or Manage Incidents.
In practice, the IT, the DevOps, and Information Systems Security (InfoSec) organizations work
jointly to ensure that organizational functions such as Patch Management are conducted
seamlessly.
Figure 15 shows how the Manage IT Infrastructure Security scope includes several Asset
Protection Management capabilities and is influenced by the Security Policy Management
Capability and Security Initiatives Management.
Zero Trust brings capabilities such as Adaptive Access Control and System and Data Asset
Protection that need to be considered.
From an information security management perspective, the Manage Asset Access Function for
data assets should include:
• The establishment of a function and capability in the security organization to:
— Develop controls, policies and procedures, and standards under the guidance of the
Risk Council
— Establish a working agreement and structure (such as Data Security Council) with the
information architecture and data governance functions to ensure that there is a clear
understanding of the lifecycle of the data element from provisioning to deprovisioning
That there is a clear understanding of its business value, and its meaning, if relevant in
an enterprise or other context.
— Establish clearly defined policies and procedures, controls, and standards so that the
policies to determine access can be clearly defined and automated
• The establishment of controls, policies and procedures, and standards to ensure clearly
defined guidance exists on what is to be done with respect to data elements with regard to
data protection
• The engagement and awareness of where the data is, and its movement, from a data loss
prevention perspective across the lifecycle of the data
Per ISO Guide 73:2009, risk management refers to the “coordinated activities to direct and control
an organization with regard to risk” [ISO 73:2009]. A risk assessment is the “overall process of
risk identification, risk analysis, and risk evaluation”, and risk analysis refers to the “process to
comprehend the nature of risk and determine the level of risk”. These are summarized in the model
shown in Figure 16.
Doing something about risk consists of implementing controls that either reduce the Loss Event
Frequency (reduce the likelihood of the Loss Scenario occurring) or reduce the Loss Magnitude
of the Loss Event once it has occurred (mitigate the severity of the loss).
Figure 18 shows the decomposition of the Loss Scenario with the Open FAIR Controls and
Categories as well as the NIST CSF color scheme.
This Chapter introduces readers to the Technical Reference Model that defines the capabilities
required for Zero Trust and realizes them into technology components. The capabilities define the
outcomes (what Zero Trust is) and the ABBs define how to build it.
This section will enable leaders and practitioners to build implementation plans and implement
the model for their organization by describing how to build Zero Trust capabilities and solutions
that are adapted to your organization’s unique business model and technical estate – most, if not
all, organizations will need to tailor Zero Trust priorities and elements to their individual situations
and requirements.
These views help business leaders and users, technology leaders and practitioners, and security
and leaders see Zero Trust from their points of view and how it applies to their roles.
Note: Throughout Chapter 5, two main colors are used to distinguish between relevant
concepts: blue is utilized for Capabilities, and green is utilized for ABBs.
As Figure 21 shows, capabilities enable doing something. In the context of the reference model,
this refers to technical capabilities, such as the ability to do access control, or log accesses. ABBs
are those logical components that implement (logically) those capabilities. Finally, SBBs are the
physical components that are how these ABBs are realized. ABBs allow visualizing the
architecture in terms of logical blocks. SBBs encompass reusing, purchasing, accessing (as in
Tooling available from vendors often provides multiple functions, allowing organizations to
rapidly implement multiple SBBs and/or ABBs with a single product (or suite of products).
6.1.1 The Zero Trust Metamodel (Adapted from the Service Oriented
Architecture (SOA) Reference Architecture Standard [C119])
Figure 22 shows the Zero Trust Metamodel.
11
Refer to Section 4.9 of the TOGAF® Standard, 10th Edition [C220].
Note: This model has been adapted from The Open Group SOA Reference Architecture, with
some modifications.
For example, the ability to determine identity for assets in a Zero Trust ecosystem is critical for
implementing adaptive asset control. Hence the ability to have a unique, reliable, digital identity
would be a Zero Trust capability. Similarly, the ability to create a secure token in lieu of the actual
sensitive data is foundational for some techniques for data centric security.
Figure 24 illustrates key Zero Trust capabilities that differentiate a Zero Trust approach from a
classic security model.
These key Zero Trust capabilities are listed below and will be described in more detail in the
subsequent subsections:
• Asset-Centricity
• Adaptive Access Control
• Digital Identity
• Asset-Centric Protection
• Asset-Centric Security Operations
• Posture Management
• Zero Trust Governance
• Security Zones
• Controls Management
Table 1 lists each Asset Centricity capability and which ABBs support it.
Table 1: Asset-Centricity Capabilities and Supporting and ABBs
Adaptive Access Control involves the ability to identify a consumer (subject), support
authentication and authorization, and implement access decisions at an individual asset level that
is informed by security risk context. Table 2 lists adaptive access control capabilities and
supporting ABBs.
Table 2: Adaptive Access Control Capabilities and Supporting ABBs
Note: Consumers are referred to as the subject, and the resource being accessed is referred to
as the asset. In practice both are assets, but in the context of access control, one asset is
acting in the role of a consumer (subject), and the other asset as a resource (object) being
accessed/consumed (often to provide a service to the consumer).
The Adaptive Access Control capability reuses the Digital Identity Binding and Asset
Management capabilities to support subject and asset identification, classification, and
management.
Data classification is used to determine one of the policies used in determining risk levels to
develop policies to apply to get access rights to resources and determine credential requirements
of the subject.
Table 3 lists each reused capability for Adaptive Access Control and which ABBs support it.
Table 3: Reused Capabilities for Adaptive Access Control
Adaptive Access Control expands the traditional two-stage process of authentication (know) and
authorization (allowed) into a three-part process that introduces explicit validation.
This provides separation and explicit definition of the outcomes of access control:
• Known (Authentication – AAC-1.1) – the subject is who they claim to be
• Trusted (Trust Validation – AAC-1.2) – the circumstances and risk factors of the access
request (and ongoing access session) are within risk tolerances/acceptability for the asset
being accessed
• Allowed (Authorization – AAC-1.3) – the subject is granted the appropriate permissions
and entitlements to the assets
Error! Reference source not found. shows that this evolution is comparable to how airport
security has evolved to meet increased threat levels where a minimum level of security is
consistently applied across all passengers beyond simply validating someone’s identity (e.g., their
government identification) and the tickets and entitlements they have bought.
Assets need to be able to support digital identity (supported by the overarching Asset Centricity
Capability), as well as entitlements. The authorization capability needs to be able to associate the
claims on the asset by a subject granted to it by an authorization provider, which obtains that
information from an asset repository. The association and enforcement of claims involves policies,
and there needs to be capabilities that define the policies, the enforcement of the policies, where
they are stored, and where the policy enforcement decision is taken. An adaptive access control
capability will also include the ability to support Agile modification of these assets (the adaptive
capability, often in the modern context involving techniques such as Machine Learning (ML)),
and the auditability and administration of these assets.
In the Digital Era, the number of participants in the ecosystem, including subjects, assets, and the
policies governing the accessing of the assets by the subject, tends to grow exponentially and
evolve rapidly. Therefore, the Zero Trust context requires the ability to support the authentication
and authorization relationship in an Agile manner to support growing ecosystems.
Adaptive Access Control also provides the ability to protect assets in an Agile manner, factoring
in multiple telemetric factors (location, business process, etc.), consumer, provided credentials
(and their classes), and resource (data) classification. Thus, this capability involves the creation of
comprehensive, rapidly evolving, and agile access control policies that might involve rapid change
in subject, asset, or the environment in which these reside, or other business, security, regulatory,
or other drivers.
Digital Identity is a key part of the ability to support agility in a Zero Trust Digital Enterprise. This
capability allows assets to support portable identity, allowing ZTAs to support the flux that exists
in relationships between organizations in the digital ecosystem in a seamless manner, without
creating a continuous set of new identities, and thus a continuous set of updates to policies and
other aspects of the ZTA.
Digital Identities support more Agile, data-centric establishment of trust, of the portability, and of
the interoperability of identity using non-repudiable attributes and credentials. Examples of
modern digital identity are the Fast Identity Online (FIDO) standard and the evolution of national
sovereign identity.12 Table 4 lists each Digital Identity capability and which ABBs support it.
All capabilities in the Asset-Centricity Capability may be reused to implement or support the
capabilities for Digital Identity.
The Adaptive Access Control capability reuses the Digital Identity Binding (AC-1.1) and other
Asset Centricity (AC-1) capabilities to support subject and asset identification, classification, and
management. These are required to enable the asset to be identified and availability protected.
Data Classification may be used to both statically and dynamically determine threat protection
levels for data assets. Thus, different data classes may be moved into different security zones, and
assets be subject to different level of protection based on data classification, intelligent threat
monitoring and dynamic risk management.
Regardless of the vector used, monetization model, or motivation, most threat actors focus on
gaining control of assets in an attack. A ZTA divides the technical estate into data and system
assets, and the security professional and the associated team must consider both data assets and
system assets in their protection design. Asset-Centric Protection is that capability that addresses
both data assets and system assets (these are the assets that manipulate that data), using Security
Zones, Adaptive Access Control, and Digital Identity.
Asset Centric Protection can be divided into protection of the two kinds of assets – data and system
assets.
• Data-Centric Protection – this capability focuses on the security of the data that matters
most to the organization
Data-centric protection enables identifying and focusing resources and security investment
on higher-value assets through their full lifecycle. It also enables identifying which assets
increase security risk without creating business value (e.g., extra copies of PII in
databases) that may be retired or tokenized. Finally, this capability allows organizations to
create flexible data protection architectures that can evolve over time, taking into
consideration the entire lifecycle of the data asset. This includes assessing the risk
associated with the data element and its use, the regulatory, business, technical, and
security implications, and the associated policies that must be created to ensure
appropriate protection and posture management. This will include data-flow analysis for
data elements.
Illustratively, for credit card numbers, the lifecycle starts from the time the number is
originated and provisioned, through different states (period that it is valid, is suspended,
lost, retention for regulatory/compliance purposes), to deprovisioning. Finally, these
would lead to a set of policies that determine the data protection regime for credit card
numbers and their use for an organization and its digital ecosystem. These policies would
be based on organizational risk appetite, regulations (e.g., Payment Card Industry (PCI)),
and other controls.
• System-Centric Protection – this capability focuses security on the assets that operate on
the data or on underlying business process, and in particular their availability
This capability uses the following capabilities Asset Centricity, Digital Identity, Adaptive
Access Control, Asset Centric Protection, and Security Zones. Core Zero Trust concepts
apply – assets are considered in terms of value and secured in Security Zones. As
expanded on in the section on Security Zones, this allows protecting assets based on
value, to address operational blast radius and compartmentalize risk, and be able to
operate in an environment of assumed breach.
Note that:
• The security responsibility for availability is focused on the intentional disruption of
services, whereas standard IT and other processes handle scenarios from natural causes,
human error, equipment failures, etc.
The remediation and scenarios may overlap between these two, but this is where Zero
Trust security focuses.
• Zero Trust is focused on all types of attack scenarios (account takeover, data corruption,
app deletion/corruption, etc.), not just network-driven scenarios
ZTAs require the ability to mitigate realized risk by limiting the time that adversaries have access
to business assets (attacker “dwell time”) with security operations. This is a natural corollary of
assuming compromise – the threat actors are assumed to gain access to assets, so the operational
security capabilities must be built to rapidly evict them. Table 6 lists each Asset Centric Security
Operations capability and which ABBs support it.
Table 6: Asset-Centric Security Operations Capabilities and Supporting ABBs
The Asset-Centric Security Operations Capability reuses the following L1 capabilities: Asset
Centricity, Asset-Centric Protection, Security Zones, and Posture Management. It is assumed that
some or all of the underlying L2 or lower capabilities of these L1 capabilities may be reused.
Asset-Centric Security Operations differ from classic security operations in the following ways:
• Asset-type specific tooling is designed to provide high quality alerts and investigation
experience, as well as proactive hunting capabilities reduces the number of false positive
alerts that cause analyst fatigue, improving overall security operations efficiency
This takes the form of EDR or XDR (ACSOP-1.1) tooling that gathers asset-specific
insights (e.g., memory and process trees from endpoints). This is often used in conjunction
with a SIEM) (ACSOP-1.7) tool and shifts some scenarios from the SIEM.
Security posture refers to an organization’s overall cybersecurity strength and how well it can
predict, prevent, and respond to ever-changing cyber threats.
This capability is critical to ensure that the organization understands and is actively reducing
potential risk to the organization.
This concept has been understood for some time: NIST 800-128 defines security posture as “the
security status of an enterprise’s networks, information, and systems based on information security
resources (e.g., people, hardware, software, policies) and capabilities in place to manage the
defense of the enterprise and to react as the situation changes. Synonymous with security status”.
Zero Trust requires a dedicated posture management capability because the concept of a “safe
network” is explicitly invalidated, so this cannot be used as a compensating control instead of
managing security posture across assets. Zero Trust also updates posture management to provide
on-demand assessment of security posture across all asset types (versus a traditional focus on
network and operating system posture) as depicted in Figure 27.
Zero Trust posture management also focuses on continuously updating posture definition and
prioritization of controls as threats, security capabilities, platforms/services, and business
priorities change.
Zero Trust Posture management explicitly expands the definition of what constitutes a
vulnerability from a functional flaw in software design or implementation into any type of
function, configuration, or operational vulnerability that allows an attacker to obtain or increase
access to the organization’s assets. These different types of vulnerabilities are depicted in Figure
28.
Figure 28: Different Types of Vulnerabilities that can Grant Attacker Control of Assets
The Posture Management capability reuses the following L1 Capabilities: Asset Centricity, Asset
Centric Protection, and Security Zones. It is assumed that some or all of the underlying L2 or
lower capabilities of these L1 capabilities may be reused.
Governance also includes the goals, principles, policies that constitute these guardrails and the
education and training required to make this actionable. Table 8 lists each Zero Trust Governance
capability and which ABBs support it.
Table 8: Zero Trust Governance Capabilities and Supporting ABBs
The Zero Trust Governance Capability reuses the Continuous Monitoring (SPM-1.1.1), Asset
Centricity (AC-1), and Data Classification (ACP-1.1.1) Capabilities.
Security Zones define secured parts of the technical estate based on various factors. They are
distinguished by:
• Protection of sets of assets of a particular business value, up to a granularity of one
(network-of-one)
• Control of the direction of the data flow (ingress/egress/routing)
• Protection of the data flowing through each zone
• Data-centric protection of the data flowing through each zone (e.g., tokenization, data
obfuscation by one-way hashing, data elimination)
• Control of access to the zone by applying least privilege and restricting the access to a
limited set of accounts, allowing zones to be architected as required by use case; the
access controls for this may include network, identity, application, data, and other types of
access controls
• Control of access by adaptive access control, allowing for multiple policies to be applied
to the access to the endpoints or zones
• Monitoring of the flow and access of data, both in real-time and through trending
Security zones simplify security design, build, and operation by providing the same or similar
security controls to this grouping of assets. Table 9 lists each Security Zones capability and which
ABBs support it.
Table 9: Security Zones Capabilities and Supporting ABBs
Note: In Table 9, the L2 capabilities associated with the L1 capabilities are assumed to be
included.
Note that the overall security partitioning in ZTAs can be considered to be a combination of
security zones as defined, division of data based on tokenization, and the setup of secure segments
based on the access controls implemented by using adaptive access control.
Zero Trust security sones incorporate support for key Zero Trust capabilities and characteristics.
Thus, data-centricity can drive the development of data-centric zones, using, for example, different
keys for different token “zones”, forming one kind of security zone. This may be seen in the
financial sector. These are network agnostic and allow organizations to reduce blast radius and
increase agility.
This capability supports blast-radius reduction, as well as complexity reduction and operational
delivery management. Modern concepts such as “Infrastructure as Code” can be used to maintain
and manage modern security zones, and provide both visibility, auditability, governance, and
runtime operational support for modern assets.
For example, in the context of hybrid cloud environments, this is usually a combination of
network/API gateway definitions, as well as access control and routing configurations driven by
the platform architecture, coupled with data-centric controls. In a hybrid cloud environment, an
environment may be setup using account controls and router and API gateway configurations to
control the routing of data, while based on business drivers, a data tokenization environment
splitting data into different token groups based on multiple factors can be set up. The first one
forms the basis for data routing, while the second one forms the basis for data sharing. The first
one uses adaptive access control to provide agility, allowing organizations to add and remove
system assets and configure the relations between them, while the second one uses data
tokenization to share data across different stakeholders such as internal APIs/capabilities, or to
clients, partners, or vendors.
13This standard does not identify or recommend a security control framework; the organization may choose to create its own or to
utilize an existing security control framework and make adaptions as required.
The Control Management capability reuses all the capabilities of Asset Centricity.
The Zero Trust Technology Reference Model applies these cross-cutting ABBs (security in
general is a cross-cutting concern) across IT, OT, and IoT environments in an organization’s
technical estate (which is comprised of all technical components in the organization).
Note: In the case of all ABBs in this document, the L2 and lower-level ABBs are subject to
evolution and change in subsequent versions of this document. Sections to be provided
in the next Snapshot version will also make the ABBs more tangible. In particular, each
lower-level ABB will be associated with scenarios, and standards for its implementation
(e.g., TLS-1.xx for the implementation of security for data in transit) will be defined.
The ABBs are arranged in two groups – those used for designing and building Zero Trust solution
architectures, and those for running and operating Zero Trust solution architectures. In some cases,
such as Posture Management, there may be ABBs for both the Design/Build group and the
Run/Operate group. These ABBs are different and address different capabilities.
These ABBs support the Zero Trust components depicted in Figure 29.
Figure 29: The Zero Trust Technology Reference Model - Design/Build View
As the ABBs are realized into SBBs and functional components (and sometimes teams) within the
organization, they will need to interact with each other, with other teams and functions, with all
the various assets in the technical estate, and with key stakeholders in IT Operations, DevOps
Teams, citizen developers, and other people within the organization. These core ABB categories
are depicted in Figure 30.
The NIST Cybersecurity Framework Functions of Identify, Protect, Detect, Respond, and Recover
illustrate the “overlay” nature of security teams. Security teams are able to read information and
context from the environment (identify/detect) freely, but work through asset owners/managers
(IT Operations) for any operations that change the environment (Prevent, Respond, Recover).
They do so by supporting the Adaptive Access Control capability and its composed lower level
capabilities of policy decisioning and enforcement. In a Zero Trust context, policy decisioning
must be centralized and consistent, and access policies must be Agile and adaptive to
accommodate rapid changes in business, technology, and security. Table 12 lists each Adaptive
Access Control capability and which ABBs support it.
The Adaptive Access Control Platform is primarily composed of the following ABBs:
• Adaptive Access Control Platform (ABB AAC-1) – Provides a policy engine and signals
that apply real time evaluation and application of organizational policy to access requests
across all types of asset (resource):
— Adaptive PIP (ABB AAC-1.1) – Is a repository of policies and support all aspects of
policy management
Policy decisioning is done using L2 policy decision point and policy enforcement point ABBs.
The policy decision point ABB can be centralized or federated but must be managed consistently.
The policy enforcement points are applied at the asset level.
Policy enforcement at an asset level allows decoupling assets from consumers/users of those assets
which are often associated with things that change frequently, such as organizational structure and
roles. Illustratively, an asset (say a microservice, an IT system, or an IOT device) can be used
within the context of an enterprise, but its owner (say an organization) might go through a merger
and acquisition, or be sold, or undergo an organization chart change in a restructuring initiative,
or have its consumers change as new partners and sales channels bring new classes of consumers.
Organizations can no longer wait long periods of time to implement these changes. The ability to
enforce these capabilities at the asset level ensures that the asset is not coupled to such structures,
and the policy decisioning ABB merely adds, deletes, or modifies policies determining the
relationship between assets and consumers.
An Adaptive Policy Decision Point ABB supports the ability to make intelligent decisions on
policy updates in order to automate the policy management process. This allows management of
the increasing complexity and volume of access control policies. Adaptive decisioning may
involve sophisticated ML approaches or simplistic decision trees. This helps incorporation of
threat intelligence and enforcement of a dynamic policy, to enable blast radius protection when an
attack is discovered, or proactively when it is expected due to threat intelligence.
The Asset Centricity and Digital Identity Platforms are reused to support identification of assets
and establishing digital identity in order for the Adaptive Access Control ABBs to be able to
operate on the assets.
The Digital Identity Platform reuses the Asset Centricity Platform ABBs.
Asset protection regimes are dependent on the asset being protected. In general, in the Zero Trust
context, asset protection of data assets involves protection of both data in use, at rest, and in flight
and the lifecycle of data including the different states that it may be in and its provisioning and
deprovisioning.
The use of this ABB has the following characteristics in a Zero Trust context:
• The application of asset protection is characterized by being performed by teams who
regularly apply these controls on these assets (IT Operations, Data Owners, DevOps
teams, etc.), helping ensure consistency across diverse types of assets
The establishment of guardrails/governance for this is a part of the establishment of the
asset protection ABB.
• Security control implementation is close to those who own the assets, empowering asset
owners and ensuring that governance and controls are applied
• The Asset-Centric Protection Platform also forms the bridge to engage IT infrastructure
and DevOps teams and provide an auditable means of control and implementation
• Data governance is responsible for data lifecycle governance of data assets, ensuring that
data-centric incorporates all stages in an asset’s lifecycle including retention,
provisioning, deprovisioning, compliance and legal requirements, classification, and other
concerns
This function also addresses any data access concerns, and integration into the business
use of the data, and the implications of any Zero Trust approaches on the business and the
data.
Table 14: Asset-Centric Protection ABBs
Zero Trust broadens the role of security operations to the full technical estate beyond the firewall
and focuses on partnership and integration with IT Operations and DevOps teams. Zero Trust also
introduces a proactive approach that includes threat hunting, continuous improvement and
automation, and red and purple team operations.
Zero Trust security operations focuses on all elements of the CIA triad including “availability” to
avoid and quickly recover from interruption of business-critical services through ransomware,
extortion, and other disruptive attacks.
This group of ABBs (Table 15) supports security operations in an asset-centric manner.
The Asset Centric Security Operations ABBs reuse the Asset Centricity Platform, and the Security
Zones Platform.
This Zero Trust function is a greatly expanded and integrated version of vulnerability scanning.
Posture management enables a proactive, holistic security approach that allows the organization
Because the posture of a complex organization is often complex and difficult to discover with a
single approach, posture management ABBs (Table 16) include both an inside-out and outside-in
view.
Table 16: Security Posture Management ABBs
The Security Posture Management Platform includes some or all of the following ABBs:
• Security Posture Management Platform (ABB SPMP-1) – Enables the organization to
discover and analyze the security posture of the organization and its exposure to attacks
and risk
— Continuous Security Posture Management Platform (ABB SPMP-1.1) – Enables
visibility into various aspects of security posture across cloud (sometimes called cloud
security posture management), on-premises, endpoints/devices, applications, network,
identity, and more
This often incorporates recommendations, best practices, and scoring as well as ML/AI
technology. This also enables individual asset owners (such as DevOps teams,
infrastructure teams, etc.) to monitor the status of their own assets so they can
continuously improve their security posture.
This ABB uses the internal vulnerability scanner (s) (ABB ACP-1.17.1) to monitor the
security status of the organization’s posture such as a CSPM tool that continuously
reports on and makes recommendations to improve security posture.
This ABB reuses the Asset Centricity Platform and the Digital Identity Platform.
Zero Trust Governance includes traditional risk, compliance, and policy functions, but they are
more dynamic (e.g., two-week sprints for policy updates rather than updates occurring every few
years).
Table 17: Zero Trust Governance ABBs
Zero Trust drives agility, adaptability, continuous learning and improvement, and reduced
complexity for a security architecture and program. Governance functions and policies are key to
enabling that while enabling the organization to meet requirements for both internal business
security assurances and external regulatory bodies.
Operational security governance provides key end-to-end services across security functions,
including risk management and policy compliance. Zero Trust introduces security architecture
and threat intelligence as governance functions to drive informed decisions across systems.
Distinguishing attributes of governance in a Zero Trust context include decisions on data
classification, data protection (including data tokenization), the establishment of security zones,
asset classification, identity management controls, and the management of policy based adaptive
access. How these capabilities interoperate and adapt to the operating environment, coupled with
the operational security governance, form the core of Zero Trust Security Governance. Continuous
Monitoring and audit on demand allows Zero Trust Governance to ensure that rapidly evolving
compliance requirements can be met.
This ABB integrates security into development of new capabilities by professional developers
(DevOps/DevSecOps teams) and Citizen Developers (low-code and no-code applications).
Figure 32 shows how Zero Trust shifts from a classic quality approval gate process that disrupts
productivity and agility to an integrated approach where security elements fit smoothly into the
Agile development process.
The Zero Trust Governance Platform reuses ABBs from the Asset Centricity, Digital Identity, and
Posture Management Platforms.
In Zero Trust terms, these ABBs include what are traditionally considered to be network zones
including tiering and micro-segmentation, as well as logical zones created through access control
Note: Technical controls must be fully supported by processes and people (training, awareness,
skills, etc.) to be effective. Real world incidents happen regularly because there are
insufficient process controls and training for how to securely support required business
workflows. One example is a fully “air-gapped” manufacturing floor that was infected
by a variant of Wannacrypt16 brought in on a hardware vendor’s maintenance laptop.
Zero Trust Security Zones can be divided into the following groups that act as L2 capabilities.
Note that a Zero Trust approach will often combine one or more of these ABBs:
• Security Zones Platform (ABB SZP-1) – Enables the organization to discretely protect
groups of highly sensitive, highly valuable, or highly fragile assets with common controls
and processes
— Identity, Endpoint, and Application-based security zone controls (ABB SZP-1.1) –
Separates the technical estate based on policies that link consumer to resource
This is implemented by the ABBs associated with Asset-Centricity, Adaptive Access
Control, Asset-Centric Protection, and Asset Availability Protection.
— Data-Centric Security Zone Controls (ABB SZP-1.2) – Allows different groupings of
data to be secured in a similar way
An example is token zones based on different keys for the same data element being
tokenized to be established. See the Data Asset Protection Platform ABB for more
details on the associated ABBs.
— Network-Centric Security Zone Controls (ABB SZP-1.3) – Creates a tiered, micro-
segmented environment
Based on solution context, would have Solution Building Blocks such as firewalls,
access control lists, and a micro-segmentation policy engine ABB.
In cloud environments, access control policies, and API gateways or service meshes
support this capability. The L2 ABBs are not itemized.
Note: Network centric controls must be supplemented by other controls to provide
strong assurances. Allowing exceptions in network controls for application and identity
protocols without providing complementary network and application-based controls
allow attackers a path to traverse a network-centric security zone boundary.
— SecOps based Zone Controls (ABB SZP-1.4) – Mitigates realized risk to assets in a
zone by limiting the time that adversaries have access to business assets
Asset-Centric Security Operations rapidly detect and remediate threats to a subset of
assets on any network, anywhere.
The ABB consists of prioritizing ABBs under Asset-Centric Security Operations Platform
(ACSOP-1) based on the sensitivity of assets in a particular Zone. This supports the Zero Trust
16 Refer to: https://www.microsoft.com/en-us/security/blog/2017/05/12/wannacrypt-ransomware-worm-targets-out-of-date-systems/.
ABBs from the Asset Centricity, Asset-Centric Posture Management, and Zero Trust Governance
Platform are reused.
During implementation, the Capabilities and ABBs defined in Chapter 5 will be used to map,
based on options (the technical environment, standards, etc.) and non-functional requirements to
SBBs which are actual components. As solution architectures are realized, the standard developers
will map the ABBs to standards and detailed implementation patterns, often determined by
scenario. These will be covered in Chapters 7-9, which will be added.
The standard developers will also be building out points that might allow compliance testing.
AI Artificial Intelligence
APP Application
CD Continuous Delivery
CI Continuous Integration
IT Information Technology
ML Machine Learning
OT Operational Technology